to kinin Pertry Ine

cyto molecular molecular

ysis contributes anal in-depth fascians cal locus

an fas ochemi the Rhodococcus production: and bi How

How the fas locus contributes to Rhodococcus fascians cytokinin production Ine Pertry ics technology net e Bio the requirements for of t and G n e illm f ers hnology in Sciences: ful ) l ec a t i reecke t 009 o r i e a (PhD V B p t 08-2 0 es lan c 2 Dr. M. Holst

: P r Dr. D. doctor rof. of yea P tor:

bmitted in u s or: ment University University

emic romo epart romot he degree of Ghent n Faculty of Scie D Acad Co-P Thesis P t FACULTY OF SCIENCES

Ghent University Faculty of Sciences

Department of Plant Biotechnology and Genetics

How the fas locus contributes to Rhodococcus fascians cytokinin production: an in-depth molecular and biochemical analysis

Ine Pertry

Thesis submitted in partial fulfillment of the requirements for the degree of Doctor (Ph.D.) in Sciences: Biotechnology

Academic year: 2008-2009

Promotor: Prof. Dr. Marcelle Holsters Co-promotor: Dr. Danny Vereecke

� � � � � � � � � � � � � � � � � � � � � � � � � � � � �

�

Onderzoek uitgevoerd in het departement Plant Sys� tems Biology (PSB) van het Vlaams Instituut voor Biotechnologie (VIB) � This work was conducted at the department Plant Systems Biology (PSB) of the Flanders Institute for Biotechnology (VIB)

� � �

Onderzoek gefinancierd � met een specialisatiebeurs van het Instituut voor de Aanmoediging van Innovatie door Wetenschap en Technologie in Vlaanderen (IWTRVlaanderen) � Research funded by a Ph.D. grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWTRVlaanderen)�

Jury Members

Promotor Prof. Dr. Marcelle Holsters (secretary) Department of Plant Systems Biology (PSB) Flanders Institute for Biotechnology (VIB) Faculty of Sciences (Ghent University) Department of Plant Biotechnology and Genetics

Co-promotor Dr. Danny Vereecke Department of Plant Systems Biology (PSB) Flanders Institute for Biotechnology (VIB) Faculty of Sciences (Ghent University) Department of Plant Biotechnology and Genetics

Promotion commission Prof. Dr. Ann Depicker (chairman) Department of Plant Systems Biology (PSB) Flanders Institute for Biotechnology (VIB) Faculty of Sciences (Ghent University) Department of Plant Biotechnology and Genetics

Prof. Dr. Danny Geelen Plant Production Faculty of Bioscience Engineering (Ghent University)

Dr. Eva Benkova Department of Plant Systems Biology (PSB) Flanders Institute for Biotechnology (VIB) Faculty of Sciences (Ghent University) Department of Plant Biotechnology and Genetics

Prof. Dr. Rosemary Loria Department of Plant Pathology and Plant-Microbe Biology Cornell University Ithaca, USA

Prof. Dr. Koen Goethals Chief Academic Administrator Ghent University

Prof. Dr. Mondher Jaziri Laboratoire de Biotechnologie Végétale Université Libre de Bruxelles (ULB)

Prof. Dr. Martin Crespi Institut des Sciences Végetales Centre National de la Recherche Scientifique (CNRS) Gif-sur-Yvette, France

Prof. Dr. Miroslav Strnad Laboratory of Growth Regulators Faculty of Sciences Palacky University, Olomouc (Czech Republic)

Table of contents

List of Frequently used Abbreviations List of Amino Acid Abbreviations

Chapter 1: 1 Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria

Hyperplasia-inducing bacteria Occurrence and biological activity of cytokinins Biosynthesis and metabolism of cytokinins tRNA as a source for cytokinins In planta functions of cytokinins

Chapter 2: 23 Rhodococcus fascians, unique amongst the hyperplasia inducing bacteria Symptom development Plant colonization by R. fascians The role of phytohormones in virulence The linear plasmid pFiD188: a mix of plasmid maintenance genes and virulence determinants The R. fascians D188 chromosome, not to be neglected in virulence contribution R. fascians dictates in planta developmental and metabolical processes Overview of the infection process

Objectives 35

Chapter 3: 39 How Rhodococcus fascians reshapes the plant: identification and modus operandi of the bacterial cytokinins Introduction Results The Arabidopsis cytokinin receptors AHK3 or AHK4 are required for symptom development. R. fascians produces a spectrum of cytokinins that are recognized by AHK3 and AHK4. Accumulation of specific R. fascians cytokinins in planta results from inadequate homeostasis. The R. fascians cytokinins act synergistically. Discussion Materials and methods

Chapter 4 69 Regulation and biochemistry of cytokinin biosynthesis in Rhodococcus fascians Introduction Results In silico analysis of the fas locus reveals new homologies and genes likely involved in fas cytokinin biosynthesis Biochemistry of the Fas proteins Novel insights in fas gene regulation Discussion Materials and methods

Chapter 5 103 A dedicated role for the Fas proteins in cytokinin production and virulence Introduction Results The spectrum of cytokinins produced by R. fascians originates from the fas operon The fas mutants have different cytokinin profiles The Fas proteins are differentially involved in virulence Perception of R. fascians cytokinins by and signal transduction through AHK4 is not maintained throughout the interaction Discussion Materials and methods

Chapter 6 127 De novo cytokinin biosynthesis in bacteria: an inheritant characteristic of phytopathogens?

Introduction Results and discussion The fas locus: an island on the linear plasmid mtr1, mtr2, fasE and fasF occurrence is not strictly correlated with cytokinin production De novo cytokinin biosynthesis mediated by a FasD-like Ipt is correlated to phytopathogenicity Concluding remarks Materials and methods

Summary and perspectives 141

Nederlandstalige samenvatting 147

References 153

Dankwoord 171

List of Frequently used Abbreviations

2MeS: 2-methylthio I: induced 2MeScZ: 2-methylthio-cis-zeatin IAA: indole-3-acetic acid 2MeScZR: 2MeScZ riboside iP: isopentenyladenine 2MeSiP: 2-methylthio-isopentenyladenine iP9G: isopentenyl-9-glucoside 2MeSiPR: 2-methylthio- iPR: isopentenyladenosine isopentenyladenosine iPRMP: isopentenyl adenosine 2MeStZ: 2-methylthio-trans-zeatin monophosphate 2MeStZR: 2MeScZ riboside IPT: isopentenyltransferase aa: amino acids Km: kanamycin AC: autoregulatory compound LDC: lysine decarboxylase add: additive LOG: lonely guy ADP: adenosine diphosphate MBP: maltose binding protein AHK: Arabidopsis histidine kinase Mtr: methyltransferase AMP: adenosine monophosphate MEP: methylerythritol ARR: Arabidopsis response regulator MVA: mevalonate Att: attenuated MUG: methylumbelliferyl-�-D-glucuronide Bp: base pairs NI: non-induced BLAST: basic local alignment tool ORF: open reading frame Cb: carbenicillin PAI: pathogenicity island CKX: cytokinine oxidase/dehydrogenase PCR: polymerase chain reaction Cm: chloramphenicol Phleo: phleomycin cZ: cis-zeatin Pyr: pyruvate

DCIP: 2,6-dichlorophenol-indophenol Q0: 2,3-dimethoxy-5-methyl-1,4-benzoquinone- DMSO: dimethylsulfoxide indophenol DMAPP: dimethylallyl diphosphate RNA: ribonucleic acid DNA: deoxynucleic acid SAM: shoot apical meristem DPI: days post infection SAM: S-adenosylmethionine DZ: dihydrozeatin SD: standard deviation Eq: equimolar SE: standard error FAD: flavine adenine dinucleotide Succ: succinate Fas: fasciation tRNA: transfer RNA FC: ferricyanide tZ: trans-zeatin GUS: �-glucoronidase uidA: �-glucoronidase HMBDP: hydroxymethylbutenyldiphosphate Z: zeatin HPLC: high presuure liquid chromatography ZR: zeatinriboside Hyp: hypervirulent

List of Amino acids abbreviations

Ala (A): Alanine Arg (R): Arginine Asn (N): asparagine Asp (D) Aspartic acid Cys (C): cysteine Gln (Q): glutamine Glu (E): glutamic acid Gly (G): glycine His (H): histidine Ile (I): isoleucine Leu (L) leucine Lys (K): lysine Met (M) : methionine Phe (F): phenylalanine Pro (P): proline Ser (S): serine Thr (T): threonine Trp (W): tryptophan Tyr (Y): tyrosine Val (V): valine

Chapter 1

Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria

Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 3

Phytohormones play a central role in regulating plant growth and development. Their relative levels define the outcome of diverse physiological processes, including differentiation and structural organisation. Plant hormones also determine the outcome of plant-pathogen interactions (Tsavkelova et al., 2006; Robert-Seilaniantz et al., 2007). Salicylic acid (SA), jasmonates (JA) and ethylene (ET) induce signalling pathways resulting in the activation of defense genes. While SA mediates mainly biotrophic resistance, JA and ET are involved in resistance to necrotrophic pathogens. However, also other plant hormones such as gibberellin, abscisic acid, brassinosteroid, auxin and cytokinin are involved in plant defense and their levels are modified upon pathogen attack (Robert-Seilaniantz et al., 2007; López et al., 2008). Many plant pathogens can interfere with these defense programs and are capable of altering the development of their host plant towards their own advantage by either producing these phytohormones themselves or by modifying the host phytohormone metabolism (Spena et al., 1992; Gaudin et al., 1994; Jameson, 2000; Tsavkelova et al., 2006; Robert-Seilaniantz et al., 2007). These virulence mechanisms have been studied extensively in gall-forming bacteria like Pseudomonas savastanoi, Agrobacterium tumefaciens, Pantoea agglomerans and Rhodococcus fascians (Jameson, 2000). Plants utilise different strategies in an attempt to counter this pathogen-induced disturbance of phytohormonal levels (Chung et al., 2008). The focus of this study is on the use of cytokinin production by the phytopathogen Rhodococcus fascians. The few described hyperplasia-inducing bacteria will be highlighted prior to discussing cytokinin metabolism in bacteria and in planta to gain insight in their mode of action. Pseudomonas savastanoi, Pantoea agglomerans, Agrobacterium tumefaciens, Agrobacterium rhizogenes, Streptomyces turgidiscabies and Rhodococus fascians are all able to induce or facilitate hyperplasia/gall formation upon infection by altering cytokinin (and auxin) levels in the host plant, although the way to accomplish this differs for each pathogen. Since R. fascians is the object of this study, it will be introduced separately in Chapter 2.

Hyperplasia-inducing bacteria

The Gram-negative bacterium Pseudomonas savastanoi pathovars nerii, fraxini and savastanoi (= P. savastanoi subsp. savastanoi) is the causal agent of knot disease on olive and other genera of the Oleaceae, and on oleander (Iacobellis et al., 1998; Pérez-Martinez et al., 2008). Disease development is defined by the formation of tissue overgrowth on different parts of infected plants. Symptoms are observed mainly at stems and branches and occasionally on leaves and fruit (see Figure 1A). The disease affects the olive oil industry mainly in Mediterranean countries reducing vegetative growth, olive yield and quality

Chapter 1 4

(Penyalver et al., 2006; Kennelly et al., 2007). P. savastanoi pv. savastanoi was reported to produce both cytokinins and auxins in culture. The amount produced varied between strains and seemed to be correlated with disease rate and the size of the developing knots (Surico et al., 1985). Whereas the production of auxin alone is sufficient to initiate knot formation, cytokinins are required for full symptom development (Iacobellis et al., 1994). For a long time auxins and cytokinins were thought to be the main virulence factors of P. savastanoi. However, an hrp defective mutant was isolated, which was unable to induce knot formation despite a normal phytohormone production (Sisto et al., 2004). In many Gram-negative phytopathogenic bacteria, hrp genes determine the ability to cause disease on hosts or to elicit a hypersensitive response in resistant plants. The hrp genes encode a type III secretion system that delivers effector proteins directly into the host cells (Tang et al., 2006). In P. savastanoi the hrp genes seem to be required for bacterial multiplication within host tissues and together with phytohormone production play a fundamental role in its pathogenesis (Sisto et al., 2004).

Pantoea agglomerans, is widespread in nature and has been associated with plants as a common epiphyte or endophyte. However, some isolates have evolved as plant pathogens. P. agglomerans pv. gypsophilae induces gall formation in gypsophila, and causes problems in gypsophila nurseries by inhibiting root development in cuttings. P. agglomerans pv. betae elicits tumorous outgrowths on both beet and gypsophila, resulting in economical losses especially in table beet industry (Barash and Manulis-Sasson, 2007; Weinthal et al., 2007). Galls are mainly induced at the crown region and have a brownish, rough and necrotic appearance (see Figure 1B) (Chalupowicz et al., 2006). Virulence is dependent on the presence of a non-conjugative pPATH plasmid containing a pathogenicity island, which carries an hrp/hrc gene cluster, virulence genes encoding effector proteins, and phytohormone biosynthetic genes (Weinthal et al., 2007). Gall initiation relies completely on type III effectors, while cytokinins and auxins exhibit a secondary role, since inactivation of the cytokinin as well as the auxin biosynthetic pathway does not prevent the formation of galls (Manulis et al., 1998). The presence of the hrp genes and a functional type III secretion system however, is absolutely required for gall initiation (Nizan et al., 1997). Cytokinins and auxins exert a significant role in virulence by controlling gall size or establishing epiphytic fitness (Lichter et al., 1995a; Barash and Manulis-Sasson, 2007). Both P. agglomerans pv. gypsophilae as P. agglomerans pv. betae produce auxin via the indole-3-acetamide (IAM) pathway and via the indole-3-pyruvate (IPyA) pathway (Barash and Manulis-Sasson, 2007). While the IAM pathway is involved in gall development and determines gall size, the IPyA pathway mediates epiphytic fitness of the bacteria (Manulis et al., 1998).

Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 5

A B C D

Figure 1: Hyperplasia induced by the discussed plant-pathogenic bacteria. (A) Knot formation on olive stems caused by P. savastanoi pv. savastanoi (Lavermicocca et al., 2002). (B) Gall induced by P. agglomerans pv. gypsophilae on Gypsophila paniculata. (C) Tumor produced by A. tumefaciens on G. paniculata (Chalupowicz et al., 2006). (D) Leafy gall formation on tobacco by the S. turgidiscabies �nos mutant (Joshi and Loria, 2007).

Streptomyces turgidiscabies is a Gram-positive phytopathogenic bacterium, which is mainly known for causing potato scab disease, resulting in significant crop losses with a high economical impact. Disease development is characterised by the production of large, erumpent lesions and symptoms are induced by the toxin thaxtomin (Loria et al., 2006). Since S. turgidiscabies is not known to induce galls in nature, it was unexpected to find a cytokinin biosynthetic pathway on a large, mobile pathogenicity island homologous and, with the exception for fasF, colinear to the 6 fas genes of R. fascians which are responsible for the development of differentiated leafy galls (for a detailed description see Chapter 2) (Kers et al., 2005). Supernatant of S. turgidiscabies exhibited cytokinin activity in callus and shoot induction assays and infection of tobacco and Arabidopsis plants with a thaxtomin-deficient mutant, resulted in the formation of leafy galls, similar to those induced upon R. fascians infection (see Figure 1D). Furthermore, consistent with its essential role in R. fascians pathology, inactivation of the ipt gene resulted in the inability to induce leafy galls. Since no gall formation was reported so far in nature, the role of this cytokinin biosynthesis pathway in potato scab disease is still speculative, although it could be responsible for the formation of the bigger, excessive lesions (Joshi and Loria, 2007).

Whereas the persistent presence of living bacteria is required for gall formation by P. savastanoi and P. agglomerans, the maintenance of crown galls is independent of the presence of its causative agent, Agrobacterium tumefaciens. This Gram-negative phytopathogen induces tumors by genetically transforming its host resulting in cytokinin and auxin overproduction. The proliferating tissue provides A. tumefaciens with a specific niche suited for its survival by producing and releasing nutritive compounds called opines (Zupan and Zambrynski, 1995; Zhu et al., 2000). A. tumefaciens has a very broad host range and

Chapter 1 6 can induce tumors at wound sites on stems and crowns, and occasionally roots of hundreds of dicotyledons and few monocotyledons (see Figure 1C) (De Cleene and De Ley, 1976). Since crown galls are a sink for water and nutrients, infection by A. tumefaciens results in a reduced crop yield, which can be significant for long-standing cultivated crops such as grape, apple and cherry (Chalupowicz et al., 2006; Escobar and Dandekar, 2003). Most of the genes mediating transformation and tumor formation are located on a Ti-plasmid. The vir regulon is responsible for the transfer of the T-DNA. Upon sensing the release of phenolic compounds, monosaccharides and the presence of acidic conditions at wound sites, the VirA/VirG two-component signalling system activates synthesis and coating of the T-strand by VirD and VirE proteins. Subsequently, the T-strand is delivered into the plant cell by a virB encoded type IV secretion system. Ultimately, the T-DNA is targeted to the nucleus where it can integrate in the plant DNA via non-homologous recombination. The transferred T-DNA, carries two sets of genes. The first set is responsible for altering cytokinin and auxin biosynthesis and sensitivity in the infected cell resulting in cell proliferation, while the second set directs the biosynthesis of amino acid and sugar phosphate derivatives called opines. Other regions are involved in replication and conjugational transfer of the Ti-plasmid and in opine uptake and catabolism (Zhu et al., 2000; Escobar and Dandekar, 2003). Agrobacterium vitis induces crown gall disease on grapevine in a similar way as A. tumefaciens, resulting in reduced growth and even death of the vines. However, in addition, A. vitis can cause via a quorum sensing system host- and tissue-specific necrosis on grape and a hypersensitive response on non-host plants, such as tobacco (Burr and Otten, 1999; Hao et al., 2005; Hao and Burr, 2006). Similar to A. tumefaciens, A. rhizogenes genetically transforms plants to produce opines. Infection results in the proliferation of transgenic roots resembling fine hairs coming through the wound sites (see Figure 2) (Srivastava and Srivastava, 2007). The genes mediating

root formation are located on the Ri plasmid. The TL-DNA carries the rol genes, which are essential for hairy root induction (Georgiev et al., 2007) and influence hormone Figure 2: Hairy root induction and secondary metabolism in transformed plants by A. rhizogenes on soybean (Kereszt et al., 2007). (Bulgakov, 2008). rolB and rolC encode �-glucosidases, which can respectively release auxins and cytokinins from their storage forms (Spena et al., 1992). Some agropine inducing strains have besides this

TL-DNA a second TR-DNA, not essential for hairy root induction, which carries auxin, mannopin and agropine biosynthesis genes. Unlike for A. tumefaciens, no cytokinin biosynthesis genes were detected on the Ri T-DNA (Casanova et al., 2005). Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 7

Occurrence and biological activity of cytokinins

Cytokinins are a group of plant hormones, which play an essential role in the regulation of plant growth and differentiation and have been recognized as central regulators of plant development. They have been shown to induce cell division, increase nutrient sink strength, delay senescence, promote outgrowth of lateral buds and inhibit cell elongation. (Sakakibara, 2006; Miyawaki et al., 2006). Since the discovery of kinetin in the fifties, several naturally occurring or synthetic compounds with cytokinin activity have been identified. Naturally occurring cytokinins are adenine derivatives with a side chain at the N6 position. They are classified into two groups depending on the structure of the side chain. Isoprenoid-type cytokinins are N6-isopentenyladenine (iP) derivatives and aromatic cytokinins carry an aromatic side chain at the N6 terminus. The side chain structure can vary in both groups by the absence or presence of hydroxyl groups and their stereo-isomeric position (see Figure 3).

N6-isopentenyladenine (iP) 2-methylthio-iP (2MeSiP) Benzylaminopurine (BAP)

Thidiazuron

trans-zeatin (tZ) cis-zeatin (cZ) Dihydrozeatin (DZ)

Figure 3: Structures of different naturally occurring and synthetic cytokinin bases (Sakakibara, 2006)

Besides iP, also hydroxylated isoprenoid-type cytokinins such as trans-zeatin (tZ), cis- zeatin (cZ) and dihydrozeatin (DZ) are common in nature. Aromatic-type cytokinins however are rare and only found in some plant species such as Arabidopsis thaliana and Populus x canadensis (Kakimoto, 2003; Tarkowska et al., 2003; Sakakibara, 2006; Miyawaki et al., 2006). Some synthetic cytokinins are structurally unrelated to these adenine derivatives, for example thidiazuron that is derived from diphenylurea (Haberer and Kieber, 2002). Cytokinins occur in different forms with different degrees of biological activity depending on the structural variations of the side chain and modifications of the adenine moiety (see Figure 4). They exist as free bases, or as their riboside or ribotide forms, and can be modified in

Chapter 1 8 several ways. Glucosylation or xylosylation results in sugar conjugates, which are less active or inactive depending on the modified position (Sakakibara, 2006, Miyawaki et al., 2006). Cytokinins can also occur as bound forms in tRNA where they can be methylthiolated at the C2 position (Esberg et al., 1999).

Figure 4: Possible modifications of the adenine moiety in cytokinins. O-glycosylation of the side chain is indicated in blue and N-glucosylation of the adenine moiety in red (Sakakibara, 2006).

In general, the free-base forms are biologically active, although cytokinin ribosides also were shown to have genuine, yet lower biological activity. The classical cytokinins, tZ and iP, display the highest activities, but also aromatic cytokinins such as benzyladenine (BA), and topolin (T) are very active. N- and O-glucosides proved to be inactive storage forms, while DZ and cZ have lower or no activities, depending on the bioassay used (Spíchal et al., 2004). However, cZ is highly abundant in some plant species such as maize, rice and chickpea and is believed to have a true hormonal function at least in some species or organs (Spíchal et al., 2004; Yonekura-Sakakibara et al., 2004). Although 2-methylthio-derivatives are only slightly less active than their unsubstituted counterparts (Matsubara, 1980), they are thought not to be of biological significance because they are believed to solely function in tRNA- mediated translation efficiency and free base forms are considered to be tRNA degradation products (Prinsen et al., 1997).

Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 9

Biosynthesis and metabolism of cytokinins

Dictyostelium discoideum, a slime mold that produces the spore germination inhibitor discadenine via iP, was the first organism in which isopentenyltransferase (Ipt) activity was demonstrated. Taya et al. (1978) showed that in this organism AMP is the direct acceptor of the isopentenyl moiety from dimethylallyl diphosphate (DMAPP) resulting in the formation of isopentenyladenosine-5’-monophosphate (iPRMP). Besides prenylating AMP, the D. discoideum Ipt also uses ADP, but not adenine, adenosine or ATP as an isopentenyl side chain acceptor (Ihara et al., 1984).

Bacterial biosynthetic isopentenyltransferases involved in gall formation

The first isopentenyltransferase gene to be characterised came from A. tumefaciens. The tmr gene resides on the T-DNA of the Ti-plasmid and tumors induced by tmr mutants formed roots and had lower cytokinin/auxin ratios compared to wild-type tumors. Furthermore the tmr gene was linked with tZ and tZ riboside (tZR) production in tumors (Akiyoshi et al., 1983). Extracts of Escherichia coli cells expressing the tmr gene were shown to have Ipt activity. The tmr gene product mediated the production of iP by transferring the isopentenyl moiety from DMAPP to AMP (Akiyoshi et al., 1984; Barry et al., 1984). Nopaline producing A. tumefaciens strains contain, besides the tmr gene, a second gene, tzs, in the vir region, which was shown to have in vitro isopentenyltransferase activity and produced iP, using DMAPP and AMP as precursors (Akiyoshi et al., 1985). Tmr as well as Tzs, only use AMP as a prenyl side chain acceptor (Sakakibara, 2006). While tmr is regulated by eukaryotic control signals and expressed in planta (Escobar and Dandekar, 2003), tzs is controlled by prokaryotic expression signals and its expression is induced by plant phenolic compounds (John and Amasino, 1988). A. rhizogenes has only one ipt gene tzs in the vir region (Regier et al., 1989; Moriguchi et al., 2000) and was shown to produce tZ (Akiyoshi et al., 1987) Based on transgenic Arabidopsis thaliana plants expressing the tmr gene, Åstot et al. (2000) suggested the existence of an alternative zeatin riboside 5’-phosphate (ZRMP) biosynthetic pathway, which directly uses a hydroxylated side chain precursor. DMAPP can be synthesised by two pathways: the mevalonate (MVA) pathway, which is present in Gram- positive bacteria and the cytoplasm of plant cells, and the MVA-independent methylerythritol phosphate (MEP) pathway, which is present in most investigated bacteria and is restricted to the chloroplasts in plants (Rohmer, 2003). Genes encoding the MVA pathway appeared to be absent in A. tumefaciens, while the genes necessary for the MEP pathway were present in its genome. An intermediate of the MEP pathway is 4-hydroxy-3methyl-2-(E)-butenyl

Chapter 1 10 diphosphate (HMBDP) and its use as a hydroxylated side chain donor was demonstrated for purified Tzs protein, thus directly synthesizing zeatin-derivatives (see Figure 5) (Krall et al., 2002). The higher biosynthetic rate of ZRMP than iPRMP in tmr transformed A. thaliana plants (Åstot et al. 2000), suggests that Tmr also accepts a precursor such as HMBDP for direct production of ZRMP (Krall et al., 2002). Indeed, in vitro studies showed that Tmr uses both DMAPP and HMBDP as side chain donors (see Figure 5). More importantly, in vivo Tmr was shown to be localized in plastids and HMBDP was the major side chain supplier (Sakakibara et al., 2005).

CH2OH

CH3

Figure 5: tZRMP biosynthesis by A. tumefaciens via an iPRMP-dependent pathway by hydroxylation of iPRMP (grey) or via an independent pathway by using HMBDP as a side chain donor.

Cytokinins are also important for P. savastanoi virulence and indeed a cytokinin biosynthetic Pseudomonas trans-zeatin producing (ptz) gene has been identified. While this gene is located mainly on plasmids called pCK for oleander isolates, it is located in the chromosome for most olive isolates (Pérez-Martinez et al., 2008). E. coli cultures expressing the ptz gene were found to secrete mainly Z and zeatinriboside (ZR), while iP and isopentenyladenosine (iPR) were present at lower levels. The expression of the tzs gene is induced by phenolic compounds, however ptz is controlled by consensus E. coli sequences and highly expressed (Powell and Morris, 1986). Since there is a striking homology between the ptz gene and the A. tumefaciens tmr and tzs genes, and bacteria expressing these genes mainly produce tZ (Powell and Morris, 1986; Akiyoshi et al., 1987), ptz might also be part of an iPRMP-independent pathway in P. savastanoi to produce tZ. The role of cytokinins in P. agglomerans disease establishment is restricted to mediating gall size. Significant amounts of secreted cytokinins were only detected in pathogenic strains Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 11 and P. agglomerans pv. gyspophilae was shown to produce mainly Z, ZR and iP (Lichter et al., 1995a). The gene responsible for cytokinin biosynthesis in this pathogen, etz, is located on the plasmid pPATH and shows high similarities to ptz, tmr and tzs and its disruption results in elimination of cytokinin production. Expression is plant-induced and seems to be regulated by a gene proceeding etz, pre-etz, although mutation of this gene resulted only in a reduction of cytokinin production (Lichter et al., 1995b; Guo et al., 2001). In silico analysis also identified a putative ipt gene in S. turgidiscabies (Kers et al., 2005). The presence of this gene is essential to induce leafy gall formation. Supernatant of S. turgidiscabies cultures shows cytokinin activity in callus initiation assays and the ipt gene is able to complement the osmosensor Sln1 yeast mutant (TM182CRE), which requires cytokinins to grow on glucose (Joshi and Loria, 2007).

Cytokinin biosynthesis in plants

Although researchers believed that in plants the cytokinin biosynthesis pathway resembled the one reported in bacteria, only few reports documented IPT biochemical activity in extracts of tobacco cytokinin-autotrophic tissue cultures (Chen and Melitz, 1979) and maize kernels (Blackwell and Horgan, 1994). It wasn’t until 2001 that nine IPT genes, named AtIPT1 to AtIPT9, with low but significant homology to the A. tumefaciens tmr gene were identified in Arabidopsis thaliana. The AtIPT and AtIPT9 proteins resembled tRNA isopentenyltransferases and AtIPT2 exhibited no IPT activity. AtIPT6 appears not to be functional in all ecotypes. Although AtIPT1 was shown to use AMP as a substrate, the Km values for AMP were much higher than those of the A. tumefaciens Ipt, indicating that AMP is not the favoured substrate. Indeed, AtIPT1 was demonstrated to efficiently prenylate ATP and ADP. Also AtIPT4 transfers an isopentenyl side chain to these substrates but not to AMP, suggesting that plant IPTs prefer ADP and ATP over AMP as a substrate (Kakimoto, 2001; Takei et al., 2001). Moreover, a phylogenetic tree demonstrated that the DMAPP:ATP/ADP isopentenyltransferase branch was composed solely of plant IPTs (see Figure 6) (Kakimoto, 2003). Since the identification of IPT genes in Arabidopsis, they were discovered in several other plant species including petunia (Zubko et al., 2002), soybean (Ye et al., 2006), pea (Tanaka et al., 2006) and Malus hupahensis (tea crab apple) (Peng et al., 2008). More importantly, biochemical characterisation of IPTs detected in hop (Sakano et al., 2004), rice (Sakamoto et al., 2006), mulberry (Abe et al., 2007) and maize (Brugière et al., 2008) showed that they more efficiently transferred an isopentenyl side chain to ADP and ATP than to AMP, confirming that plant Ipts indeed act as DMAPP:ATP/ADP isopentenyltransferases.

Chapter 1 12

Figure 6: Phylogenetic tree of plant and bacterial isopentenyltransferases (Kakimoto, 2003). Ipts cluster into three different branches according to their substrate specificity: 1. DMAPP: ATP/ADP isopentenyltransferases (red): plant IPTs 2. DMAPP:tRNA isopentenyltransferases (blue): widespread in different organisms 3. DMAPP: AMP isopentenyltransferases (green): bacterial Ipts

In Arabidopsis, two pathways were suggested for the production of zeatin: an iPRMP- independent (Åstot et al. 2000) and an iPRMP-dependent pathway (Takei et al., 2004). When zeatin is formed via iP, this occurs by hydroxylation of the prenyl side chain mediated by two cytochrome P450 monooxygenases, CYP735A1 and CYP735A2 (see Figure 7). These enzymes were shown to preferentially hydroxylate iP nucleotides over iPR and iP, hereby producing mainly tZ. The cis-isomer was produced at a neglectable level (Takei et al., 2004). tZ can be further converted to dihydrozeatin by a reductase, which was detected in Phaseolus vulgaris seeds (Mok and Mok, 2001). In the iPRMP-independent pathway zeatin would be produced by directly transferring a hydroxylated side chain precursor to adenine Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 13 nucleotides (Åstot et al. 2000) and indeed, AtIPT7 was able to use HMBDP as well as DMAPP as side chain donor (Takei et al., 2003). AtIPT1, 3, 5 seemed to be correlated with iP as well as tZ production and they are localised to plastids, where the MEP pathway, delivering HMBDP besides DMAPP as a side chain donor, is active. This suggested that these proteins could directly produce tZ (Kasahara et al., 2004; Miyawaki et al., 2006). However, their overexpression resulted predominantly in iP accumulation and labelling experiments in wild type and AtIPT1 overexpressing Arabidopsis plants indicated that DMAPP is the main side chain donor for tZ production in planta (Sun et al., 2003; Sakakibara et al., 2005). Mevastatin, an MVA pathway inhibitor, reduced iPRMP-independent tZ biosynthesis, suggesting that the side chain precursor is delivered via the MVA pathway (Åstot et al. 2000). Since an extensive fraction of the cZ side chain is delivered via the MVA pathway (Kasahara et al., 2004), another hypothesis proposed that iPRMP-independent tZ biosynthesis in planta would proceed via cZ derivatives (Sakakibara et al., 2005). The produced cZ could then subsequently be converted to tZ mediated by a cis-trans isomerase (see Figure 7), although to date this enzyme was only detected in P. vulgaris (Bassil et al., 1993). Finally, the biologically active cytokinin forms are thought to be the nucleobases. There are two distinct pathways to convert cytokinin nucleotides to their free bases (see Figure 7). The first one involves a two-step reaction in which the cytokinin nucleotides are subsequently dephosphorylated and deribosylated. Partially purified wheat germ extracts exhibited high 5’- ribonucleotide phosphohydrolase activity towards AMP and iPMP, while ADP and ATP were dephosphorylated at a lower rate (Chen and Kristopeit, 1981a) and adenosine nucleosidase activity towards adenosine and iPR (Chen and Kristopeit, 1981b). However, to date, the corresponding genes have not been identified yet. More recently, a second free-base releasing pathway was discovered in rice, where the LONELY GUY (LOG) gene encodes a protein with cytokinin-specific phosphoribohydrolase activity, which results in the direct liberation of free-base forms from cytokinin monophosphate nucleotides. A homologous LOG gene family was also detected in the Arabidopsis genome (Kurakawa et al., 2007).

Cytokinin metabolism

Different mechanisms can control the levels of biologically active cytokinins. First of all, the biosynthesis rate plays an important role. Secondly, cytokinins can be (ir)reversibly inactivated by degradation or conversion to storage or ribonucleotide forms. The latter shares enzymes of the purine salvage pathway. Although an important role of adenine phosphoribosyltransferases (APRTs) is the recycling of adenine into adenylate, two

Chapter 1 14

APRTs were characterised in Arabidopsis, APT2 and APT3, with high affinities for cytokinins (Allen et al., 2002). Cytokinins are irreversibly degraded by cytokinin oxidase/dehydrogenases (CKXs) to adenine/adenosine and its corresponding side chain aldehyde (see Figure 7) (Werner et al., 2006). These products are not released directly, but are the result of a hydrolytic cleavage of an intermediate cytokinin imine product (Popelková et al., 2006). For a long time, the reaction was thought to require oxygen and the enzyme was classified as an copper amine oxidase (Rinaldi and Comandini, 1999). However, for wheat CKX it was demonstrated that the reaction did not require oxygen and did not produce hydrogen peroxide, indicating that the enzyme acts as a dehydrogenase (Galuszka et al., 2001). Later on, the enzyme appeared to exhibit a dual action. When oxygen is used, the enzyme has a low substrate specificity and the rate of the reaction is low. In contrast, the reaction proceeds at high rates and strict specificities for iP and analogous cytokinins when specific electron acceptors such as DCIP and certain quinones are used (Frébortová et al., 2004). In general, it can be stated that iP, its riboside and tZ are good substrates for CKX enzymes, ZR and cZ are worse substrates and although aromatic cytokinins can be degraded the reaction proceeds at very low rates (Bilyeu et al., 2001; Galuszka et al., 2001; Frébortová et al., 2004; Galuszka et al., 2004; Popelková et al., 2006). DZ, cZR and cytokinin metabolites such as zeatin-9-glucoside, and O-glucosides appear not to be degraded at all (Bilyeu et al., 2001; Frébortová et al., 2004; Schmülling et al., 2003). Although cytokinin nucleotides also are considered not to be degraded, they are potential substrates of some AtCKXs (Galuszka et al., 2007). To date, putative or characterised CKX genes were reported in several plants such as Arabidopsis (7 CKX genes), rice (11 CKX homologues), barley, wheat, maize, orchid, cotton, Medicago truncatula and pea (Bilyeu et al., 2001; Yang et al., 2003; Galuszka et al., 2004; Werner et al., 2006; Wang et al., 2008). Ckx activity was also detected in two non-plant species: the slime mold Dictyostelium discoideum (Armstrong and Firtel, 1989) and, albeit extremely low, in Saccharomyces cerevisiae (Schmülling et al., 2003). Moreover, putative CKXs were also found in Nostoc, a cyanobacterium (Werner et al., 2006) and two Actinomycetes, R. fascians (Crespi et al., 1994) and S. turgidiscabies (Joshi and Loria, 2007).

Cytokinin storage forms are produced by glycosylation of the hydroxyl group of the side chain of tZ, DZ and cZ or of the adenine moiety at the N3, N7 and N9 position (see Figure 7) (Sakakibara, 2006). Two enzymes mediating O-glycosylation have been identified: O- glucosyltransferase (OGT) and O-xylosyltransferase (OXT). The latter was isolated from bean (P. vulgaris) and can convert tZ and DZ, but not cZ and ZR to the corresponding xylosyl derivatives (Martin et al., 1999b). OGTs also seem to act stereo specific. In maize, two OGT genes were identified, which are specific for cZ (Martin et al., 2001; Veach et al., 2003), while Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 15 an OGT isolated from lima bean (P. lunatis) modified tZ and DZ to O-glucosylzeatin (ZOG), but not cZ or ZR (Martin et al., 1999a). The OGT isolated from lima bean can also glycosylate the aromatic cytokinin meta-topolin (mT) as a substrate and one of the cis-stereo specific OGTs can glycosylate ortho-topolin (oT) and hydroxylated thidiazuron (Mok et al., 2005). O-glucosylation can be reversed, the storage forms are released to active bases by �- glucosidases (Brzobohatý et al., 1993). Two enzymes were identified in Arabidopsis that glucosylate the adenine moiety at the N7 or N9 position resulting in the formation of N7 or N9 glucosides (7G or 9G) (Hou et al., 2004). Unlike O-glucosylation, N-glucosylation is thought to be irreversible (Sakakibara, 2006).

Figure 7: Model of cytokinin metabolism pathways in Arabidopsis (modified from Sakakibara, 2006). AK: adenosine kinase; APRT: adenine phosphoribosyltransferase, �Glc: �-glucosidase; ZOGT: zeatin O-glucosyltransferase; CK-N-GT: cytokinin N-glucosyltransferase. 1: phosphatase dephosphorylating iPRTP and iPRDP, 2: 5’-ribonucleotide phosphohydrolase, 3: adenosine nucleosidase, 4: purine nucleoside phosphorylase, 5: zeatin cis-trans isomerase, 6: zeatin reductase, 7: hypothetical cytokinin cis-hydroxylase . iPRDP: isopentenyladenosine-5’-diphosphate, iPRTP: isopentenyladenosine-5’-triphosphate, tZRDP: tZR 5’diphosphate, tZRTP: tZR 5’-triphosphate, cZR: cZ riboside, cZRMP: cZR 5’monophosphate.

Chapter 1 16

tRNA as a source for cytokinins

Since many years the role of tRNA as a cytokinin source has been discussed. It was shown for A. tumefaciens that the tRNA makes a only a small, yet significant contribution to the amount of secreted iP (Gray et al., 1996). Furthermore, calculations of the tRNA turnover rates in plants indicated that they were too slow to account for the measured amounts of cytokinins (Klämbt et al., 1984; Klämbt, 1992). Frequently, tRNAs of plants, animals and most bacteria, but not Archaebacteria, that read codons starting with U, contain cytokinin residues (Golovko et al., 2002). They are always located at position 37 (see Figure 8) and modify tRNA translation efficiency (Prinsen et al., 1997). In animals, iP is the only detected cytokinin in tRNA. In plants, cytokinins are detected in cytoplasmic, plastid and mitochondrial tRNA (Golovko et al., 2002) and cZ predominates. For bacterial tRNAs, 2-methylthio- adenosine (2MeSiPR) seems to be the major cytokinin component (Prinsen et al., 1997). To Figure 8: Cytokinin residues in tRNA are date, methylthiolated nucleosides have been found located immediately adjacent the anticodon loop at position 37 (Modified to be exclusively synthesised in tRNA (Anton et al., from http://www.chemistryexplained.com/ 2008). Pr-Ro/Protein-Synthesis.html). In plants as well as in bacteria, IPTs mediating prenylation of tRNA have been identified. In Arabidopsis, AtIPT2 and AtIPT9 show homology to tRNA IPTs (Kakimoto, 2001). AtIPT2 was shown to exert tRNA IPT activity and complemented tRNA modification in a yeast mutant strain in which MOD5, the sole tRNA IPT gene in yeast (Dihanich et al., 1987), was disrupted. Although cZ is the major cytokinin component in tRNA, no cZ was produced in the complemented yeast cells (Golovko et al., 2003) suggesting that AtIPT2 can not directly transfer a hydroxylated chain, but that an additional, still unidentified, enzyme is required for the hydroxylation modification. An AtIPT2 mutant showed reduced levels of cZ, cZR and cZRMP and similarly, in an AtIPT9 mutant the latter cytokinin metabolites were reduced. In the AtIPT2,9 double mutant both these cis-derivatives were undetectable, which strengthens the idea that cZ-type cytokinins are tRNA degradation products (Miyawaki et al., 2006). Nevertheless, the high levels of cis-isomers in some organs or plant species, such as maize and chickpea, suggest that tRNA breakdown is insufficient to account for the detected amounts (Veach et al., 2003). Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 17

In bacteria, several tRNA modifications have been described (see Figure 9). Firstly, tRNA iPR is synthesized by MiaA, a tRNA Ipt that was identified in bacteria such as E. coli (Caillet and Droogmans, 1988), Salmonella typhimurium (Ericson and Björk, 1986) and A. tumefaciens (Gray et al., 1996). Subsequently, MiaB, identified in Thermotoga maritime (Pierrel et al., 2003), E. coli and S. typhimurium (Esberg et al., 1999), thiolates and methylates tRNA iPR at position 2 of the base moiety. While S-adenosylmethionine (SAM) delivers the methyl group, the sulfur atom is provided by the enzyme itself (Pierrel et al., 2004). Finally, in S. typhimurium, MiaE hydroxylates 2MeSiPR in tRNA to form 2-methylthio- cis-zeatinriboside (2MeScZR) (Persson and Björk, 1993; Mathevon et al., 2007).

Figure 9: Action of the different Mia enzymes on tRNA in S. typhimurium (Mathevon et al., 2007).

In planta functions of cytokinins

Cytokinins are important plant growth regulators and direct many physiological processes, such as leaf senescence, stress tolerance, vascular differentiation, chloroplast biosynthesis, nutrient balance, and root, shoot and inflorescence growth and branching (Müller and Sheen, 2007) and promote the outgrowth of axillary buds (Chatfield et al., 2000). Cytokinins positively regulate the shoot apical meristem (SAM). CKX overexpressing plants have a lower cytokinin content and a smaller SAM due to a slower proliferation and a tardy formation of leaf primordia, resulting in retarded shoot development (Werner et al., 2001; Werner et al., 2003). Moreover, it was shown that cytokinins mediate cell cycle progression (Laureys et al., 1998; Dobrev et al., 2002), underlining their stimulating effect on cell proliferation. In contrast to their stimulating action on the SAM, cytokinins inhibit the root meristem by controlling the cell differentiation rate (Kyozuka, 2007). While addition of cytokinins decreases meristem size, depletion increases root meristem size as demonstrated by the development of larger roots and more lateral roots by cytokinin biosynthesis mutants (Dello Ioio et al., 2007), CKX overexpression plants (Werner et al., 2001; Werner et al., 2003) and cytokinin receptor mutants (Riefler et al., 2006). Cytokinin overexpression plants further

Chapter 1 18 indicated or suggested that cytokinins stimulate leaf growth via cell division, control vascular development by stimulating cambial activity and influence reproductive development and embryo maturation (Werner et al., 2003). Cytokinin receptor mutants showed that cytokinins delay leaf senescence by retaining chlorophyll, regulate seed size and germination and control cytokinin metabolism (Riefler et al., 2006).

Cytokinin perception and signalling

The cytokinin signal transduction pathway is a phosphorelay resembling bacterial two- component signalling systems (see Figure 10) (Heyl and Schmülling, 2003). The first identified cytokinin receptor gene was the CYTOKININ RESPONSE 1/ARABIDOPSIS HISTIDINE KINASE 4 (CRE1/AHK4). Mutants displayed a cytokinin-insensitive phenotype and expression in yeast confirmed a cytokinin-dependent phosphorelay activation (Inoue et al., 2001; Ueguchi et al., 2001). Besides AHK4, also AHK2 and AHK3 function as cytokinin receptors (Yamada et al., 2001; Higuchi et al., 2004; Nishimura et al., 2004; Romanov et al., 2006). While AHK4 is mainly expressed in the root, AHK2 and AHK3 are expressed throughout the plant, albeit to a different degree (Higuchi et al., 2004). AHK4 is highly specific and sensitive towards iP and tZ. AHK3 on the other hand, seems to have a broader, albeit less sensitive, ligand specificity (Spíchal et al., 2004). Cytokinin receptor mutant analyses underlined their role in plant development and cytokinin-dependent processes. Although ahk3 mutants develop a slightly smaller rosette, in general, single receptor mutants and, ahk2ahk4 and ahk3ahk4 double mutants are not visibly affected, suggesting that these receptors function redundantly. Root growth of ahk4 mutants is not inhibited by cytokinins, indicating that AHK4 is the main root receptor (Inoue et al., 2001). ahk2ahk3 double mutants on the other hand, develop smaller leaves, but have a normal root system, indicating that these receptors function mainly in the shoot (Higuchi et al., 2004; Riefler et al., 2006). Besides these three genuine cytokinin receptors, in Arabidopsis, another gene CYTOKININ INDEPENDENT-1 (CKI-1) is involved in cytokinin signalling, although it is currently unclear if CKI-1 functions as a cytokinin receptor or as a positive regulator of signal transduction (Glover et al., 2008). The cytokinin receptors are thought to mediate extracellular cytokinin signals and are essential for normal growth en fertility. However, the fact that triple receptor mutants can establish a basic plant body suggests the existence of a cytokinin receptor- independent system, which is hypothesised to regulate the cell cycle autonomously via an intracellular cytokinin pool (Higuchi et al., 2004; Nishimura et al., 2004; Riefler et al., 2006). The current model on cytokinin perception and signalling is presented in Figure 10. Upon cytokinin perception via an extracellular cytokinin-binding CHASE (cyclases/histidine kinases Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 19 associated sensory extracellular) domain, a conserved histidine residue in the cytoplasmic kinase domain is autophosphorylated (Heyl et al., 2007). Subsequently, the phosphoryl group is transferred via the receptor receiver domain to a phosphotransfer protein (AHP) (Mahönen et al., 2006). In Arabidopsis, five AHPs, which positively regulate cytokinin signalling, and AHP6, a pseudo-AHP, which lacks the crucial histidine phosphorylation site and negatively regulates cytokinin signalling by preventing the phosphotransfer to AHPs, were identified (Suzuki et al., 2000; Hutchison et al., 2006; To and Kieber, 2007).

Figure 10: Model of cytokinin perception and signalling in plants (To and Kieber, 2007). See text for details.

Upon phosphorylation, the AHPs translocate to the nucleus where they transfer the phosphorylgroup to type-A or type-B response regulators (RRs) (To and Kieber, 2007). Eleven type B RRs were identified in Arabidopsis (ARRs). They mediate cytokinin signalling by activating transcription of cytokinin target genes (Mason et al., 2005; Argyros et al., 2008), including A-type ARRs, which are stabilised and activated by phosphorylation. Ten A-type ARRs were identified and eight of them negatively regulate cytokinin signalling, although they also exhibit roles in other physiological processes such as SAM regulation and light

Chapter 1 20 responses (To et al., 2007). In addition, cytokinin response factors (CRFs) have been identified, which act in a cytokinin-dependent manner, but independently of A- and B-type ARRS. Upon cytokinin perception and signalling via the AHKs and AHPs, they are translocated to the nucleus, where they control transcription of cytokinin target genes. Unlike type-B ARRs, they do not control type A ARR transcription (Rashotte et al., 2006; Müller and Sheen, 2007).

Localised cytokinin biosynthesis, degradation and transport regulate plant growth and development

Cytokinin biosynthesis, degradation and signalling occurs throughout the plant and is not necessarily co-localised (Miyawaki et al., 2004; Mason et al., 2004; Werner et al., 2006). Cytokinin biosynthesis occurs in a wide range of organs and cell types: AtIPT1 is expressed in root xylem founder cells, lateral buds, ovules and immature seeds; AtIPT4 and AtIPT8 in developing seeds; AtIPT5 in the root cap and primordia, lateral buds, young inflorescences and fruit abscission zones; AtIPT3 in the phloem throughout the plant and AtIPT7 in root elongation endodermis cells and trichomes on young leaves. Although tRNA IPTs are thought to account only for a minor contribution to the cytokinin pool of the plant, AtIPT2 and AtIPT9 are expressed overall and the highest expression levels are observed in proliferating tissues (Miyawaki et al., 2004). CKX expression on the other hand, is restricted to dividing cells of young tissues: AtCKX1 and 2 are expressed in the shoot apex, AtCKX4 in young leaves and stomatal precursor cells, AtCKX5 also in young leaves and the root meristem procambial region, and AtCKX6 in the vascular system (Werner et al., 2006). Moreover, different CKXs occur in different subcellar compartments: CKX1 and 3 are targeted to the vacuole and CKX2, 4, 5 and 6 are predicted to function in the apoplast (Werner et al., 2003). The differential spatial expression of these genes suggests that cytokinins can act both as local or long-distance signals and can be degraded or perceived at sites differing from their sites of origin (Hirose et al., 2008). This is supported by the finding that IPTs are expressed in the phloem and xylem (Miyawaki et al., 2004) and the detection of cytokinins in the xylem and phloem sap (Sakakibara, 2006). Cytokinin transport can be mediated by common purine transporters, which can also transport adenine. Purine permeases (PUP) are a family of 15 integral membrane proteins which are suggested to be involved in import of purines, secondary metabolites such as caffeine and nicotine, and cytokinins, although cytokinin transport activity was only detected for PUP1 and PUP2 (Gilissen et al., 2000; Bürkle et al., 2003). Equilibrative nucleoside transporters (ENTs) might be responsible for cytokinin translocation. At least one of the four homologues in rice (AtENT2) and one of the Cytokinins, naturally occurring plant hormones used as virulence factors by phytopathogenic bacteria 21 eight homologues in Arabidopsis (AtENT6), mediate iPR transport (Hirose et al., 2005; Hirose et al., 2008). Despite the fact that no IPTs were found to be expressed in the SAM (Miyawaki et al., 2004), cytokinins play an important role in SAM maintenance by sustaining cell proliferation (Kyozuka, 2007). Since IPTs are not expressed in the SAM, cytokinins are thought to be imported into this tissue and function as long-distance signals (Miyawaki et al., 2004). However, the LOG gene, is in rice also required for normal SAM development and is expressed specifically in shoot meristem tips suggesting a local cell-specific cytokinin action (Kurakawa et al., 2007).

In conclusion, we can state that cytokinins and more specifically their tightly controlled levels, are very important plant growth regulators which determine the outcome of diverse plant developmental processes. However, several plant pathogens have evolved strategies to disrupt this hormonal balance by transforming the plant cells to phytohormone overproduction factories or by producing these phytohormones themselves and alter the physiology of the plant to their own advantage.

Chapter 2

Rhodococcus fascians, unique amongst the hyperplasia inducing bacteria

Rhodococcus fascians , unique amongst the hyperplasia inducing bacteria 25

At the beginning of the previous century fasciations were described on pea as short, fleshy stems carrying misshaped leaves (Brown, 1927). Few years later the bacteria responsible for these distortions were isolated (Tilford, 1936). After a long-term discussion on its taxonomic position, the bacterium was finally classified as Rhodococcus fascians (Goodfellow, 1984). R. fascians is an aerobic, pleiomorphic, non-motile Gram-positive Actinomycete, which forms moist-cream to orange colonies and grows as branched hyphae, which rapidly fall apart in rods and cocci (LeChevalier, 1986). R. fascians was classified within the Mycobacteriaceae based on rRNA analysis, cell wall composition, the presence of tuberculostearic and mycolic acids, and the high GC content (61-67%). R. fascians is the only known phytopathogenic Rhodococcus species (Luehrsen et al., 1989; Bell et al., 1998; Goethals et al., 2001). Strain D188, which is studied during this thesis, carries a circular chromosome, a circular plasmid, involved in cadmium resistance but not in pathogenesis, and a conjugative linear plasmid pFi, essential for virulence (Desomer et al., 1988; Crespi et al., 1992; Pisabarro et al., 1998).

Symptom development

R. fascians possesses a very broad host range, which encompasses up to 40 different plant families, 87 genera and 122 taxa that mostly belong to the dicotyledones and include woody and herbaceous plants. R. fascians is widespread around the world and has been isolated in Northern Europe, the Middle East, Asia, Australia, New Zealand, Canada, Mexico, and at least 19 states of the United States. Since R. fascians infection results in the development of malformed shoots and flowers or decreased flower production, effects were mainly noticed in the ornamentals industry, where it is a persistent problem that can lead to enormous economical losses (Depuydt et al., 2006; Putnam and Miller, 2007). Morphological changes induced upon infection depend on the plant species (Lacey, 1939), the age of the plant (Roussaux, 1965), bacterial growth conditions (Faivre-Amiot, 1967), the bacterial strain (Eason et al., 1995), and the inoculation method (Vereecke et al., 2000) and vary from leaf deformation, over the formation of witches’ brooms to leafy gall development. Although wounding of the plant is not required, it will enhance symptom development. Despite the fact that R. fascians can act on all cell types competent to divide, meristems and buds seem the most sensitive to infection. The development of the root system is generally inhibited in tobacco upon infection (Vereecke et al., 2000). Leaf deformation in tobacco is characterised by swelling of the petioles and veins, caused by enlarged parenchyma cells and secondary growth of the vascular tissues, and

Chapter 2 26 wrinkling of the lamina, on witch green islands occur (Vereecke et al., 2000). A bunch of fleshy stems with misshapen leaves on the crown of pea are typical for a witches’ broom (Roussaux, 1975) and infected plants are dwarfed and produce less flowers (Lacey, 1936). Infection of Arabidopsis thaliana leads to the development of an overall bushy and stunted plant with reduced chlorophyll content, anthocyanin accumulation, newly-formed small, serrated leaves, and multiple rosettes and inflorescences (see Figure 1B) (Vereecke et al., 2000; de O Manes et al., 2004; Depuydt et al., 2008). Tobacco seedling infection results in a complete growth inhibition (Vereecke et al., 2000), whereas infection of A. thaliana seedlings only results in growth retardation (de O Manes et al., 2004). The most severe symptom is the development of leafy galls, which consist of multiple shoot primordia that are inhibited in their outgrowth (see Figure 1A). Leafy galls are formed by the activation of existing axillary meristems and the development of new meristems from differentiated tissue (Goethals et al., 2001). They are not necrotic, but are characterized by delayed senescence, they remain long green after the host plant is senescing. The presence of the bacteria is absolutely necessary for the persistence of the galls. When bacterial activity is eliminated, shoots will grow out, form roots and develop into normal plants (Vereecke et al., 2000).

A B C D

Figure 1: Symptoms induced by R. fascians on different hosts. A. Leafy gall formation on tobacco. B. A. thaliana infection results in bushy and stunted plants C. Petunia sp. D. Atropa belladonna

Plant colonization by R. fascians

R. fascians is a well-adapted epiphyte and the bacterium mainly colonizes aerial plant parts. The bacteria occur on both sides of the leafs and locate mainly on leaf edges, the stem and crown are colonised to a lesser extent (Cornelis et al., 2001; Vereecke et al., 2003). The presence of R. fascians on a host plant does not necessarily lead to the development of symptoms, indicating a strict regulation of the bacterial virulence signals. Triggered by Rhodococcus fascians , unique amongst the hyperplasia inducing bacteria 27 unknown environmental conditions the bacteria will invade the plant tissues. Although stomata are colonized, they are not used as preferred entryways (Cornelis et al., 2001). Instead, the bacteria penetrate directly through the cuticle and epidermis via the formation of ingression sites, which is not accompanied by extensive cell necrosis (Vereecke et al., 2003). From one week post infection on, bacteria can be observed inside the plant tissue. They are mainly located in intercellular spaces and occur throughout the tissues, although fewer bacteria reside in the deeper cell layers (Cornelis et al., 2001). In contrast to other phytopathogens, such as Agrobacterium tumefaciens, Pseudomonas savastanoi and Pantoea agglomerans, R. fascians has not been detected in vascular tissues (Goethals et al., 2001). The role of the different bacterial populations on and in plants is not completely resolved. The epiphytic bacteria appear to initiate symptom development in response to environmental conditions, while the endophytic population is apparently responsible for the persistence of the leafy galls and profits from the newly formed niche for survival (Maes et al., 2001; Vereecke et al., 2002; Depuydt et al., 2008b).

The role of phytohormones in virulence

The fine tuned balance of plant growth regulators plays a key role in growth and development of healthy plants. Many plant-associated bacteria however are capable of influencing their eukaryotic hosts by either producing phytohormones themselves or by modulating phytohormone production by the plant (Tsavkelova et al., 2006; Robert- Seilaniantz et al., 2007). In contrast to the undifferentiated galls induced by P. agglomerans, P. savastanoi or A. tumefaciens, phytohormone production by R. fascians provokes the formation of differentiated leafy galls. Many of the symptoms triggered by R. fascians, including abundant shoot proliferation and inhibition of root development, are typical cytokinin effects and addition of cytokinins has been found to mimic some of the symptoms (Thimann and Sachs, 1966; Oduro and Munnecke, 1975; Depuydt et al., 2008). However, despite the detection of several cytokinins in the culture supernatant of R. fascians (Klämbt et al., 1966; Helgeson and Leonard, 1966; Rathbone and Hall, 1972; Scarbrough et al., 1973; Armstrong et al., 1976; Murai et al., 1980; Eason et al., 1996), to date, no clear positive correlation could be made between the amount of secreted cytokinins, the virulence of the producing strain and the kind of cytokinins detected in infected plants. Moreover, contradictory results in planta make it impossible to link cytokinins to symptom development (Goethals et al., 2001; Depuydt et al., 2008; Murai et al., 1980; Balázs and Sziráki, 1974; Eason et al., 1996; Crespi et al., 1992; de O Manes et

Chapter 2 28 al., 2001; Galis et al., 2005). It has been suggested that R. fascians infection would stimulate cytokinin production by the plant (Goethals et al., 2001). However, recent studies showed that this is not the case because in planta cytokinin biosynthesis is downregulated upon R. fascians infection (Depuydt et al., 2008). Moreover, expression of the plant CKX genes and consequently cytokinin degradation was upregulated (Galis et al., 2005; Depuydt et al., 2008). Most likely this reflects an attempt of the plant to maintain homeostatic hormone levels and indicates that the bacteria constantly deliver cytokinins into the plant tissues, thereby inducing and maintaining symptoms. More clarity about the role of cytokinins was provided by the characterization of an essential isopentenyltransferase (ipt) gene located on the linear plasmid pFiD188 and its strict correlation with virulence in a high number of isolates tested (Crespi et al., 1992; Stange et al., 1996). Moreover, this gene was also identified on a chromosomal pathogenicity island of S. turgidiscabies, the only other phytopathogen known to be capable of inducing differentiated leafy galls (Kers et al., 2005; Joshi and Loria, 2007). In Chapter 3, a more detailed introduction is given on what has been reported on R. fascians cytokinin production when we describe the results on the identification and modus operandi of these bacterial morphogens. In combination with the secretion of cytokinins, some R. fascians strains appear to be able to degrade auxin (Roussaux, 1975; Kemp, 1978), which would contribute to the increased cytokinin/auxin ratio and support symptom development. However, some aspects of symptom development like cell swelling, secondary differentiation of vascular tissue and induction of lateral root initiation are typical auxin effects and indeed infected plants contain significantly more indole acetic acid (IAA) than control plants (Vereecke et al., 2000). Moreover, it was shown that R. fascians strain D188 can produce and secrete IAA (Vandeputte et al., 2005). Besides a possible role in the initiation of symptoms by suppressing plant defense (Robert-Seilaniantz et al., 2007; Depuydt et al., 2008b), auxin could also function in the epiphytic survival of the bacteria (Goethals et al., 2001).

The linear plasmid pFiD188: a mix of plasmid maintenance genes and virulence determinants

R. fascians carries a conjugative linear plasmid with is a mix of genes essential for pathogenicity and involved in plasmid maintenance. Comparative analysis of pFiD188 with the linear plasmids of the soil bacteria R. erythropolis BD2 and PR4, and Rhodococcus sp. strain RHA1 (see Figure 2) revealed the presence of four colinear regions (R1, R2, R3 and R6) and more importantly, three regions which are unique for pFiD188 (U1, U2 and U3) (Francis et al., 2007). Rhodococcus fascians , unique amongst the hyperplasia inducing bacteria 29

Figure 2: Comparative analysis of the linear plasmids of R. fascians (pFiD188), R. erythropolis BD2 (pBD2), R. erythropolis PR4 (pREL1) and Rhodococcus sp. strain RHA1 (pRHL2) revealed the presence of three unique regions for pFiD188 (Francis et al., 2007).

The conservation of the R regions implies that they are involved in plasmid maintenance. Region R2 is involved in the conjugal transfer of the linear plasmid (Pertry et al., unpublished data), while region R1 appears to be involved in the replication of the linear plasmid. No role can be attributed to the regions R3 and R6 because they carry genes, which encode hypothetical proteins (Francis et al., 2007). The unique regions are likely involved in virulence and/or the interaction with the plant host. Region U2 harbours an nrp locus, which putatively encodes a nonribosomal peptide synthetase producing a pentapeptide, of which the role remains to be unravelled. The stk (serine threonine kinase) locus resides in the region U3 and is hypothesized to be involved in the early steps of the interaction (Francis et al., 2007). The unique region U1 hosts four loci of which three (fas, att and hyp) were identified via random insertion mutagenesis as either essential for or involved in virulence (Crespi et al., 1992). The att (attenuated) locus comprises nine genes (see Figure 3). While attA, attB and attH seem to be involved in arginine biosynthesis, the attD, attE and attF gene products show homology to �-lactamring forming proteins. Mutations in the att locus lead to an attenuated virulence, characterised by the development of less or smaller leafy galls. Hence, the att locus is responsible for the production of an unidentified autoregulatory compound, which is actively exported via AttX and is proposed to have an antibiotic-like structure (Maes et al., 2001). Based on the available data, a model for its function was proposed. The att compound

Chapter 2 30 is produced constitutively at a low level until a certain threshold is reached and a positive autoregulatory loop is activated via AttR, a LysR-type transcriptional regulator. Activation of this loop leads directly or indirectly to the expression of other virulence-associated genes, such as the fas genes. The att molecules also appear to be involved in the penetration of plant tissues, since the att mutant, in contrast to the wild-type strain, requires wounding to induce mild symptoms on tobacco (Maes et al., 2001). The attenuated phenotype induced by att mutants is assumed to be the result of a delayed fas gene expression eventually combined with a reduced endophytic colonisation (Cornelis et al., 2002).

Figure 3: Schematic representation of the att locus in the region U1 on the linear plasmid pFiD188 of R. fascians (modified from Maes et al., 2001).

The hyp (hypervirulent) locus encodes a protein homologous to the DEAD-box family of RNA helicases, which play an important role in mRNA stability, and initiation of translation. Mutation in this locus lead to the development of bigger leafy galls, suggesting that it is involved in controlling the virulence signals by modulating virulence gene expression (Crespi et al., 1992; Temmerman, 2000). In between the att and the hyp locus a gene cluster is located which is putatively involved in the biosynthesis of Gram-positive A-factor signalling molecules (our unpublished data). A more detailed view on the possible role of the hyp locus and A-factors in the R. fascians-plant interaction is given in Chapter 3, when we discuss the results on fas gene regulation. The fas (fasciation) locus consists of the fas operon and the fasR gene, which encodes an AraC-type regulator, and was shown to be essential for virulence (Crespi et al., 1992; Crespi et al., 1994; Temmerman et al., 2000). The fas operon harbours six genes ORF1- ORF6, currently renamed fasA-fasF, of which the gene products make up the molecular machinery responsible for the biosynthesis of a cytokinin that is directly involved in virulence (Crespi et al., 1994). The importance of this operon is reflected by its strict regulation, both at the transcriptional and translational level which is influenced by multiple environmental factors (Temmerman et al., 2000; Maes et al., 2001; Cornelis et al., 2002). Moreover, the complete fas operon was identified in the only other known differentiated gall inducing bacterium, S. turgidiscabies (Kers et al., 2005; Joshi and Loria, 2007). The fas locus is the Rhodococcus fascians , unique amongst the hyperplasia inducing bacteria 31 topic of this thesis and therefore a more detailed introduction is given in Chapter 3, when we discuss its regulation and biochemistry.

The R. fascians D188 chromosome, not to be neglected in virulence contribution

Besides encoding the key pathogenicity determinants, the linear plasmid is also suggested to fine-tune the chromosome, which seems to play a significant role in the adaptation towards an endophytic lifestyle. A chromosomal mutant was isolated which showed severely deficient virulence. The mutation was located to the vicA gene of the vic locus (virulence in the chromosome) gene, which encodes a malate synthase required for the glyoxylate shunt of the Krebs cycle. Nutrient catabolism leads to the formation of glyoxylate, a toxic metabolite, which is titrated out by its conversion to malate via malate synthases. A functional vicA gene is thus absolutely necessary for the removal of this toxic component and for endophytic survival. Although the vicA gene resides on the chromosome, its expression was shown to be regulated by the linear plasmid (Vereecke et al., 2002). As mentioned earlier it has been demonstrated that R. fascians can produce and secrete indole-3-acetic acid (IAA) and that plants carrying leafy galls contain more IAA than control plants. Analysis of IAA production by a plasmid-free strain showed that the biosynthetic genes reside on the bacterial chromosome. Although the genes are yet to be determined, biochemical data indicate that their kinetics and regulation is influenced by the presence of the linear plasmid (Vandeputte et al., 2005).

R. fascians dictates in planta developmental and metabolical processes

Infection of A. thaliana by R. fascians results in an altered leaf shape: leaf margins are strongly serrated and the laminae fail to expand. In general, newly formed leaves are much stronger affected than leaves already present at the time of infection. The formation of these misshaped leaves is correlated with the upregulation or ectopic expression of the class I KNOTTED-like homeobox (KNOX) family transcription factors (Depuydt et al., 2008), which play a crucial role in meristem function and maintenance (Hake et al., 2004). Moreover, albeit that knox mutants were responsive to R. fascians, they developed no or milder serrations (Depuydt et al., 2008). KNOX transcription factors are believed to reduce gibberillic acid and increase cytokinin levels, resulting in hormone ratios favouring meristematic tissue identity

Chapter 2 32

(Hake et al., 2004). However, R. fascians infection results in decreased cytokinin levels and increased gibberillic acid (GA) levels, implying that KNOX gene expression is rather a consequence of the altered hormone balance than its cause. The constant delivery of bacterial cytokinins, reflected by the induced plant cytokinin homeostatic mechanism, might trigger KNOX gene expression, while the high gibberillic acid levels may result from the amplification of young GA-producing tissue (Depuydt et al., 2008). Although ectopic KNOX expression is responsible for the formation of serrated leaf margins, it does not account for the leaves failure to expand. The small lamina size is not a consequence of a reduced cell number, instead R. fascians stimulates the plant cells to divide and prevents them to proceed into endoreduplication-driven cell expansion (Depuydt et al., 2008a). R. fascians was shown to accelerate cell cycle progression of tobacco BY-2 cells (Vandeputte et al., 2007). In tobacco, the first steps towards leafy gall formation comprise de novo cell division of outer cortical stem cells, which is accompanied by CYCD3;2 expression indicating a re-entry of these differentiated cells into the cell cycle (de O Manes et al., 2001). Similarly, in A. thaliana, CYCA2;1, CYCB1;1 and CDKA;1, which are linked to active cell division, were induced upon infection (Vereecke et al., 2000). CYCD3 and CDKB1;1 mediate, respectively, the G1-to-S and G2-to-M cell cycle transition and are thought to be direct targets of the bacterial virulence signals resulting in a fast progression through mitosis. While in normal developing Arabidopsis plants the expression of several tested cell cycle genes decreased over time, their expression in infected tissue continuously increased, which keeps the cells in a proliferative state and prevents differentiation and maturation (Depuydt et al., 2008a). The proliferation of young tissues is believed to generate a specific niche for the bacteria, providing them with plenty of nutrients. Indeed, a thorough transcript and metabolic profiling revealed an altered metabolic state of infected A. thaliana plants favourable for the bacteria (Depuydt et al., 2008b). Tryptophan accumulation might stimulate bacterial auxin biosynthesis, whereas decreased arginine and ornithine and increased succinate and pyruvate levels are conditions promoting bacterial virulence gene expression (Temmerman et al., 2000; Maes et al., 2001; Vandeputte et al., 2005; Depuydt et al., 2008b). Moreover, molecules involved in plant development, such as polyamines and trehalose, accumulate and might amplify the cytokinin triggered processes. Amino acid levels increased over time and might serve as a nitrogen source for the bacteria. The metabolic changes furthermore included an increase in carbohydrates, such as maltose and glucose, which could play a role in the energy demanding niche establishment process, but are also good carbon sources for the bacteria (Depuydt et al., 2008b).

Rhodococcus fascians , unique amongst the hyperplasia inducing bacteria 33

Overview of the infection process

Epiphytic stage Endophytic stage Niche development

� Bacteria reach the plant � nrp encoded production of a � Continued expression of fas surface (stk?) phytotoxin facilitates bacterial genes leads to a continuous entry via ingression sites supply of cytokinins

� Epiphytic colonies are � Full induction of att gene � Formation of shoot embedded in a protective expression results in a high meristems and primordial slime layer threshold concentration of autoregulatory compound

� Auxin production (iaa) � Activation of fas gene � The resulting leafy gall facilitates nutrient acquisition expression and cytokinin exhibits an altered physiology production

� Expression of the att genes � Dedifferentiation of cortical � Adaptation of the bacterial leads to a gradual build up of cells and induction of cell metabolism to the new niche the concentration of the division via the vic locus autoregulatory compound

Figure 4: Schematic overview of the R. fascians-plant interaction (Francis et al., 2007)

A schematic overview of the interaction is given in Figure 4. R. fascians can easily survive in the soil, but is also a well-adapted epiphyte. A crucial first step in conquering the plant is reaching the plant surface. Since R. fascians is non-motile, most likely the bacteria reach the plant by chance via rain drops, insect vectors, etc. or possibly a more active process is encoded by the stk locus which is hypothesized to mediate a plant-directed growth. As soon as contact is established, epiphytic colonies embedded in a protective slime layer may start secreting auxin, thereby suppressing plant defense and/or stimulating nutrient release. When bacteria are triggered to engage in an endophytic life phase, a phytotoxin release putatively produced by the non-ribosomal peptide synthetase and/or an att-related function are responsible for the formation of ingression sites through which they enter the plant tissues. Initially, the att molecule is synthesised at a low constitutive level until a certain threshold is reached and a positive autoregulatory loop is activated leading to full induction of att gene expression and subsequent activation of fas gene expression. Although att gene

Chapter 2 34 expression is limited to bacteria located on the plant surface and during early stages of infection, expression of the fas genes continues throughout infection and coincides with sites of meristem initiation and proliferation. A constant production of fas-dependent cytokinins is absolutely necessary for the maintenance of the symptoms. Through the persistent activation of the plant cell cycle, hyperplasia are induced in which the metabolome is different from non-symptomatic tissues. In this way a specific niche is created that is efficiently metabolically colonized by R. fascians (Vereecke et al., 2000; Maes et al., 2001; Cornelis et al., 2002; Depuydt et al., 2008a; Depuydt et al., 2008b; Francis et al., 2007).

Objectives

Objectives 37

Based on the shooty phenotype and the presence of an ipt gene on the linear virulence plasmid of R. fascians D188, the role for cytokinins in the pathology had been anticipated for a long time. Subsequent studies identified and characterized the fas operon as a key genetic determinant of virulence and likely cytokinin biosynthesis. Nevertheless, many aspects concerning regulation of fas gene expression, Fas protein function, and, importantly, the encoded cytokinin biosynthetic pathway and the identity of the produced morphogens remained to be uncovered. Therefore, the main objectives of this research were to identify the bacterial cytokinins responsible for the R. fascians pathology, to unravel how they exerted their function, and to elucidate the role of the fas locus and its expression in their production. We set out to identify and quantify the in vitro produced cytokinins of two near-isogenic R. fascians strains only differing in the presence of the linear plasmid, grown under the previously determined optimised conditions for fas gene expression, using an activity-based approach combined with advanced chromatographic techniques. By assessing the biological activity in different bioassays, the in planta stability, and the involvement of the plant’s cytokinin perception machinery in the recognition of the identified molecules, we hope to unravel different aspects on their mode of action. A thorough in silico analysis is anticipated to provide useful hints on the function of each Fas protein in the production of the cytokinin(s) and should allow to postulate a working model for the biosynthetic pathway. The predicted enzymatic activities will be tested by expressing the Fas proteins in a heterologous system and using the purified proteins in in vitro assays. We will generate mutants in each of the fas genes making up a different functional module and analyse the cytokinin spectrum of their supernatants. With this biochemical approach we hope to tighten the working model. To comprehend the in planta function of the different functional modules, the virulence of the generated mutants will be evaluated. To obtain clues on how fas-dependent cytokinin production might be controlled, the in vitro and in planta expression patterns will be determined and the role of postulated regulatory mechanisms will be analysed. Finally, we wonder how common Ipt- and/or fas-dependent cytokinin production is in the bacterial world. To address this question we will analyse the phylogenetic relationship of homologues of the different Fas proteins. Following these approaches it is our ambition to uncover the enigma of the R. fascians pathology or at least to shed new light on the virulence strategies of this fascinating bacterium.

Chapter 3

�

How Rhodococcus fascians reshapes the plant: identification and modus operandi of the bacterial cytokinins

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

Parts of this chapter are published as� “Identification of Rhodococcus fascians cytokinins and their modus operandi to reshape the plant” (PNAS, 2009, 106(3):929-34). � 1 1 1 � Pertry I , Václavíková K , Depuydt S , Galuszka P, Spíchal L, Temmerman W, Stes E, Schmülling T, Kakimoto T, Van Montagu MCE, Strnad M, Holsters M, Tarkowski P, and Vereecke D (1: equal contribution).�

Identification and modus operandi of the bacterial cytokinins 41

Introduction

The fine-tuned balance of plant regulators plays a key role in growth and development of plants. Many plant-associated bacteria can influence their hosts either by modulating phytohormone production or by producing phytohormones themselves. The main advantages for the bacteria are increased nutrient release, suppression of defense, and/or establishment of disease (Tsavkelova et al., 2006; Robert-Seilaniantz et al., 2007). The latter is most outspoken for hyperplasia-inducing bacteria such as Pantoea agglomerans and Pseudomonas savastanoi that secrete high amounts of cytokinins and auxins to facilitate or initiate gall development (Barash and Manulis-Sasson, 2007; Sisto et al., 2004), and Agrobacterium tumefaciens that genetically transforms plant cells thereby converting them into cytokinin and auxin (and opine) factories (Zhu et al., 2000). In contrast to the undifferentiated galls induced by the bacteria mentioned above, the Actinomycete Rhodococcus fascians, that shares persistence strategies with the closely related human pathogen Mycobacterium tuberculosis (Vereecke et al., 2002), provokes the formation of differentiated leafy galls consisting of numerous shoot primordia that are inhibited in further outgrowth (Goethals et al., 2001). The shooty symptoms that can partially be mimicked by exogenous addition of cytokinins (Thimann and Sachs, 1966; Depuydt et al., 2008), initiated more than 40 years ago experiments to characterize the bacterial cytokinins. Analyses of culture supernatant of different non-isogenic virulent and avirulent R. fascians strains grown under rich culture conditions identified 11 different cytokinins: isopentenyladenine (iP), methylaminopurine (Helgeson and Leonard, 1966), cis-zeatin (cZ) (Scarbrough et al., 1973), 2-methylthio-cis-zeatin (2MeScZ), isopentenyladenosine (iPR) (Armstrong et al., 1976), 2- methylthio-isopentenyladenine (2MeSiP), cis-zeatinriboside (cZR), 2-methylthio-cis- zeatinriboside (2MeScZR) (Murai et al., 1980), trans-zeatin (tZ), trans-zeatinriboside (tZR), dihydrozeatin (DZ) and dihydrozeatinriboside (DZR) (Eason et al., 1996). With the exception of iP, the primary source for these cytokinins was assumed to be tRNA (Murai et al., 1980). Despite the fact that these cytokinins are produced by R. fascians, contradictory results currently make it impossible to positively correlate the amount of secreted cytokinins, the virulence of the producing strain and their presence in infected plants. Although an elevated level of cytokinins was detected in infected Pelargonium zonale stems, other reports could not observe a significant difference in cytokinins in infected plants. Even more, in Arabidopsis thaliana and pea a lower total cytokinin content was measured upon infection (Goethals et al., 2001; Depuydt et al., 2008; Murai et al., 1980; Balázs and Sziráki, 1974; Eason et al., 1996; Crespi et al., 1992; de O Manes et al., 2001; Galis et al., 2005). Additional

Chapter 3 42

complications arise in assigning a role for cytokinins in the R. fascians pathology when the secreted amounts are taken into account. Even when a weak correlation could be made between bacterial cytokinins and virulence, none of the researchers were convinced that the measured levels could account for the observed effect of the bacteria on plant development (Goethals et al., 2001; Eason et al., 1996). Moreover, the levels of some cytokinins detected in cultures of A. tumefaciens and P. savastanoi were up to 1000-fold higher than those reported for virulent R. fascians strains (Eason et al., 1996; Akiyoshi et al., 1987). Why then is only R. fascians able to induce shoots at the infection site? The role of cytokinins in R. fascians-induced symptoms was strongly supported by the identification of an isopentenyltransferase (ipt) gene on a linear virulence plasmid (Crespi et al., 1992) and its strict correlation with virulence (Stange et al., 1996). Moreover, Streptomyces turgidiscabies, the only other bacterial species known to induce differentiated leafy galls, differs from other scab-causing Streptomycetes by the unique acquisition of a R. fascians-like ipt gene (Joshi and Loria, 2007). The ipt gene of strain D188 was expressed only under very specific conditions that likely reflect the presence of the host, and importantly, are very different from those used in the past for identification and quantification of the bacterial cytokinins (Crespi et al., 1992; Temmerman et al., 2000). Here, we studied the role of the bacterial cytokinins in the R. fascians pathology by assessing the involvement of the cytokinin perception machinery of A. thaliana in symptom development. The cytokinin profile of two near-isogenic R. fascians strains, only differing in the presence of the linear plasmid was reassessed and the biological activity of the identified cytokinins, their in planta stability, and their capacity to activate the cytokinin signaling cascade in Arabidopsis was determined. Based on the results, we postulate a model on the mode of action of the bacterial cytokinins during symptomatology.

Identification and modus operandi of the bacterial cytokinins 43

Results

The Arabidopsis cytokinin receptors AHK3 or AHK4 are required for symptom development.

Figure 1: Response of single, double and triple cytokinin receptor mutants upon mock inoculation or infection with R. fascians D188 at 17 dpi. Responsiveness is evidenced by activated axillary meristems and serrated leaf margins.

The typically stunted and bushy phenotype that develops in Arabidopsis upon R. fascians infection is correlated with a strong cytokinin response, including the activation of homeostatic mechanisms and expression of A-type response regulators (Depuydt et al., 2008). To assess their role in disease, several cytokinin receptor mutants (Inoue et al., 2001; Higuchi et al., 2004) were infected and their response evaluated over time. Single receptor mutants in ahk2, ahk3 and ahk4 (see Figure 1), and the double mutants ahk2ahk4 and ahk2ahk3 developed regular

Chapter 3 44

R. fascians disease symptoms with regular kinetics (see Figure 1). In contrast, the ahk3ahk4 mutant, in which AHK2 is the only functional cytokinin receptor, and the triple knockout mutant did not respond to the pathogen (see Figure 1), demonstrating that AHK3 or AHK4, but not AHK2, are necessary for symptom development. We evaluated the spatio-temporal dynamics of the expression of the cytokinin perception machinery via histochemical analysis of infected AHK3:GUS and AHK4:GUS lines in a time- course experiment. The AHK3 promoter activity was very high, both in the complete shoot of mock-inoculated controls and in symptomatic tissues (see Figure 2A). However, the AHK4 expression was activated from 4 days post infection (dpi) onward, both in the preformed leaves and in induced symptomatic tissues, indicating a broadening of the expression domain. The induction level was stronger in the youngest leaves, which may be correlated with the colonization degree (see Figure 2B). Together, these data prove the crucial role of the bacterial cytokinins in the pathology.

A

B

Figure 2: Histochemical analysis of AHK3:GUS (A) and AHK4:GUS (B) lines upon mock-inoculation (control) or infection with R. fascians D188. Expression patterns were visualized at 4 or 14 days post treatment.

R. fascians produces a spectrum of cytokinins that are recognized by AHK3 and AHK4.

To identify the bacterial cytokinins responsible for virulence, the cytokinin profiles were determined of the pathogenic strain D188 and its non-pathogenic plasmid-free derivative D188- 5, grown overnight in defined medium under control and optimized conditions for virulence gene expression (Temmerman et al., 2000). Whereas the levels of O- and 9-glucoside derivatives, benzylaminopurine, p-, m- and o-topolin and DZ were below detection limit, the Identification and modus operandi of the bacterial cytokinins 45

cytokinin ribosides had the same profile as their bases, but were produced at concentrations below 1 nM (see Figure 3).

1.2 0.016 D188-5 NI D188-5 I D188 NI D188 I 0.014 1 0.012 0.8 0.01

nM 0.6

nM 0.008 0.4 0.006 0.004 0.2 0.002 0 0 iPR 2MeSiPR cZR 2MeScZR tZR Figure 3: Cytokinin riboside profiles of the supernatants of near-isogenic R. fascians strains D188 and D188-5 grown under non-inducing (NI) and inducing (I) conditions for virulence gene expression. Succinate was added to the medium as a carbon source. Error bars represent SDs (n=3).

Interestingly, both strains produced the same cytokinin spectrum (see Figures 3 and 4), implying that R. fascians D188 virulence is a matter of surpassing cytokinin threshold levels rather than producing specialized molecules. Indeed, when focusing on the cytokinin bases, significantly higher levels of 2MeScZ (2-fold), iP (5-fold), and cZ (2-fold) were measured for strain D188 than for D188-5, 2MeSiP levels were independent of the linear plasmid. The concentration of tZ was 100-fold lower than that of the other cytokinins, but a 2-fold increase was observed in the presence of the linear plasmid (see Figure 4). Although 2-methylthio-trans- zeatin (2MeStZ) was detected in D188 supernatants only, the levels were very close to the detection limit (data not shown). However, the level of cytokinins did not clearly differ when bacteria were grown under non- inducing or inducing conditions for virulence gene expression, suggesting that in vitro growth conditions are not adequate to support efficient cytokinin production. Indeed, bacterial IAA production is often only measurable when an excess of the precursor tryptophan is added to the medium. This is also the case for R. fascians IAA production (Vandeputte et al., 2005). The isopentenyl side chain donor, required for the production of isoprenoid-type cytokinins can be delivered by the non-mevalonate pathway via pyruvate (Hunter, 2007). Interestingly besides succinate, which was added for the inductions of the cultures for the cytokinin quantifications, pyruvate also has a strong enhancing effect on fas gene expression

(Temmerman et al., 2000).

Chapter 3 46

Figure 4: Cytokinin profile of the supernatant of near-isogenic R. fascians strains D188 and D188-5 grown under non-inducing (NI) and inducing (I) conditions for virulence gene expression. Succinate was added to the medium as a carbon source. Error bars represent SDs (n=3).

Since pyruvate is a potential precursor for the isopentenyl side chain, we reasoned that its presence in the induction medium could possibly activate the biosynthetic machinery by installing a feed forward loop. Therefore, cytokinin quantifications were repeated under inducing and non-inducing conditions using pyruvate instead of succinate. Although the cytokinin profiles obtained under these conditions were different from those with succinate (see Figures 4 and 5), again no significant induction could be observed. Moreover, the differential accumulation of tZ, 2MeScZ and cZ in the presence of the virulence plasmid, was no longer observed. iP levels on the other hand, were higher and still correlated with pFiD188 and 2MeSiP was produced approximately at the same concentration. Also under these conditions, the cytokinin ribosides had the same profile as their bases and were produced in much lower levels (data not shown). Altogether, these data underline that there is a discrepancy between the effect of carbon sources on fas gene expression and cytokinin production and emphasizes the importance of environmental conditions on the functioning of the cytokinin biosynthetic machinery. Identification and modus operandi of the bacterial cytokinins 47

7 0.035 D188-5 NI D188-5 I D188 NI D188 I 6 0.03 5 0.025 4 0.02 M nM 3 n 0.015 2 0.01 1 0.005 0 0

iP 2MeSiP cZ 2MeScZ tZ

Figure 5: Cytokinin profiles of the supernatants of near-isogenic R. fascians strains D188 and D188-5 grown under non-inducing (NI) and inducing (I) conditions for virulence gene expression. Pyruvate was added to the medium as a carbon source. Error bars represent SDs (n=3).

Interestingly, in the literature, little biological activity is attributed to cZ and 2MeScZ (Matsubara, 1980). Therefore, the ability of AHK3 and AHK4 to recognize the six cytokinin bases produced by R. fascians was tested in an receptor-binding assay, using a system where AHK3 or AHK4 are functionally expressed in Escherichia coli and their output signal activates an E. coli signaling pathway coupled to the lacZ reporter gene (Spíchal et al., 2004). The dose response curves show that 2MeScZ, 2MeSiP and cZ bind both cytokinin receptors, although less efficiently than the other cytokinins (see Figure 6). tZ and 2MeStZ are very effectively perceived by both receptors and iP seems to be better recognized by AHK4.

Figure 6: Dose-dependency of cytokinin-induced �-galactosidase activity of AHK3 and AHK4 expressed in E. coli. The �-galactosidase activity of non-induced strains (control) is indicated by the dotted line. Error bars represent SDs (n=3). Next, we monitored the phenotypic alterations in the double receptor mutants triggered by

Chapter 3 48

adding different cytokinins to the medium. Fourteen days old mutant plants were transferred to medium containing 10 µM of the cytokinin of interest and grown for 10 days. As shown in Figure 7A the response was generally weak when only AHK2 was active, except for iP and 2MeSiP. When only AHK3 or AHK4 was present, similarly strong responses were observed. When the double receptor mutants were transferred to medium containing different concentrations of cytokinin mixes, the presence of AHK2 alone was sufficient to trigger responses, but mainly at higher concentrations (see Figure 7B).

2MeStZ

Figure 7: Phenotypic response of the cytokinin receptor double mutants towards (A) individual cytokinin treatment at 10 �M and (B) different concentrations of equimolar cytokinin mixes. In B, ahk2ahk3 is not shown but the response was similar to that of ahk2ahk4; eq4 (iP, 2MeSiP, cZ, and 2MeScZ); eq6 (iP, 2MeSiP, cZ, 2MeScZ, tZ and 2MeStZ).

Accumulation of specific R. fascians cytokinins in planta results from inadequate homeostasis.

If the cytokinins identified above were biologically important during R. fascians infection, they would be expected to be stable in planta and to occur in infected tissues. Previous studies have demonstrated that upon infection with R. fascians the overall cytokinin content decreased due to rapidly activated cytokinin homeostatic mechanisms (Depuydt et al., 2008; Galis et al., 2005). To isolate and identify cytokinins from infected Arabidopsis and tobacco plant tissue, plants were harvested at 8 and 4 weeks post infection respectively, and cytokinins were Identification and modus operandi of the bacterial cytokinins 49

extracted from 1 to 10 grams of material as described in Materials and Methods and fractionated via HPLC under acidic conditions. Fractions were collected every minute, redissolved in 10 µL of 80% methanol and tested in a yeast bioassay (for a detailed description of this assay, see Chapter 5). However, no bioactive fractions were detected, indicating that cytokinin concentrations in infected plants are very low.

Figure 8: Cytokinin profiles in A. thaliana Col-0 shoots at different time points during the interaction with R. fascians. The respective cytokinin is indicated above each graph; mock-inoculated with water (dashed line, white circle), upon infection with strain D188 (full line, black circle). Error bars represent SDs (n=3).

Because 2MeS-type cytokinins had never been considered and quantitative analysis is a more sensitive technique to measure cytokinins, we analyzed the cytokinin content of infected Arabidopsis shoot material prior to visible symptoms (2 dpi), at the onset of tissue deformation (7 dpi), and after fully established disease (35 dpi), and compared it to mock-inoculated controls. Generally, the cytokinin concentrations were low, confirming previous obtained results. Whereas tZ levels were lower in infected tissue already at 2 dpi, the other cytokinin bases were initially more abundant upon infection (see Figure 8). As this timing coincides with

Chapter 3 50

the onset of bacterial virulence gene expression in planta (Cornelis et al., 2002; Chapter 6) and the cytokinin biosynthetic machinery of the plant is repressed early upon infection (Depuydt et al., 2008), the observed cytokinin accumulation can directly be attributed to bacterial synthesis. With time, iP, 2MeSiP, and 2MeStZ levels decreased, but 2MeScZ and especially cZ accumulated (see Figure 8), resulting eventually in a 30-fold and 3-fold increase respectively.

The build-up of cZ and 2MeScZ might result from the inability of cytokinin oxidase/dehydrogenases (CKXs) to degrade these cytokinins. In its endophytic life phase, R. fascians resides in the apoplast (de O Manes et al., 2004; Cornelis et al., 2001), and hence, the apoplastic enzymes CKX2, CKX4, and CKX6 of Arabidopsis (Werner et al., 2003) are probably most important for cytokinin homeostasis during the R. fascians pathology. Interestingly, these 3 CKXs happen to be the strongest upregulated upon R. fascians infection (Depuydt et al., 2008). CKX1 and CKX3 are targeted to the vacuole (Werner et al., 2003) and are likely of less importance for the R. fascians-Arabidopsis interaction. The sequence of CKX5 suggests an apoplastic localization, although this has not been experimentally proven. The phenotype of plants overexpressing CKX5 is similar to that of plants overexpressing vacuolar CKX genes (Schmülling T, personal communication) and the substrate specificity of CKX5 resembles that of the vacuolar CKX enzymes (see Table 2). Partially purified recombinant CKX1-6 proteins were used in in vitro CKX enzyme assays with the six R. fascians cytokinin bases and their ribosides as substrates. Ferricyanide (FC, dehydrogenase activity) or oxygen (O2, oxidase activity) were used as electron acceptors, at the physiological pH 6.0. The capacity of each CKX enzyme to degrade the tested cytokinins was expressed as absolute enzymatic activities and as relative values compared to their ability to degrade iP (see Tables 1 - 4). Strikingly, although the substrate specificities varied, cZ and 2MeScZ were the worst cytokinin base substrates for all three apoplastic CKXs (see Table 1), which probably contributes to the accumulation of these cytokinins in infected plants. The vacuolar CKX enzymes showed a very different substrate specificity and degraded the cis- and 2MeS-derivatives quite well (see Table 2). Similar specificity patterns, were observed towards the cytokinin ribosides (see Tables 3 and 4). Comparable effects on CKX activities were monitored while using 2,6-dichlorophenolindophenol (DCIP) or 2,3-dimethoxy-5-methyl-1,4- benzoquinone (Q0) and similar phenolic compounds, which are thought to be natural electron acceptors for the cytokinin degradation machinery in vivo, as electron acceptors (data not shown).

Identification and modus operandi of the bacterial cytokinins 51

CKX2 CKX4 CKX6 FC O2 FC O2 FC O2 pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % iP 250 100 2.7 100 47.4 100 0.29 100 750 100 6.1 100 2MeSiP 72 29 0.5 19 22.9 48 0.15 52 403 54 1.3 21 cZ 23 9 - - 11.2 24 - - 262 35 - - 2MeScZ 5 2 - - 4.5 9 - - 130 17 - - tZ 75 30 3.0 111 39.4 83 0.28 97 224 30 3.4 56 2MeStZ 130 52 0.4 15 8.1 17 0.11 38 318 42 4.2 69

Table 1: Substrate specificity of the apoplastic CKX enzymes of Arabidopsis in dehydrogenase (ferricyanide: FC) and oxygen (O2) mode at pH 6.0 given as absolute (pkat/mg) and relative activities. All data represent mean values of at least two biological replicates. Deviations between replicates did not exceed 10%; -: oxidative degradation of cis-derivatives can not be determined (see Materials and Methods).

CKX1 CKX3 CKX5 FC O2 FC O2 FC O2 pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % iP 2332 100 8.2 100 466 100 15.7 100 500 100 36.3 100 2MeSiP 3597 154 5.8 71 508 109 13.8 88 529 106 14.8 42 cZ 3306 142 - - 0.5 0.1 - - 191 38 - - 2MeScZ 3742 160 - - 5.5 1.2 - - 437 87 - - tZ 3178 136 41.6 507 93 20 3.1 20 158 32 39.9 111 2MeStZ 3838 165 23.7 289 513 110 5.0 32 482 96 79 219

Table 2: Substrate specificity of the vacuolar CKX enzymes and CKX5 of Arabidopsis in dehydrogenase (ferricyanide: FC) and oxygen (O2) mode at pH 6.0 given as absolute (pkat/mg) and relative activities. All data represent mean values of at least two biological replicates. Deviations between replicates did not exceed 10%; -: oxidative degradation of cis-derivatives can not be determined (see Materials and Methods).

CKX2 CKX4 CKX6 FC O2 FC O2 FC O2 pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % iPR 150 60 2.82 104 35.1 74 0.32 110 871 116 10.2 167 2MeSiPR 83 33 0.70 26 43.6 92 0.09 31 524 70 2.6 43 cZR 1 0.4 - - 3.1 7 - - 219 29 - - 2MeScZR 2 0.8 - - 1.6 3 - - 116 15 - - tZR 138 55 3.28 122 52.6 111 0.31 107 215 29 5.4 89 2MeStZR 65 26 0.29 11 15.8 33 0.07 24 171 23 3.8 62

Table 3: Substrate specificity of the apoplastic CKX enzymes of Arabidopsis in dehydrogenase (ferricyanide: FC) and oxygen (O2) mode at pH 6.0 given as absolute (pkat/mg) and relative activities. All data represent mean values of at least two biological replicates. Deviations between replicates did not exceed 10%; -: oxidative degradation of cis-derivatives can not be determined (see Materials and Methods).

Chapter 3 52

CKX1 CKX3 CKX5 FC O2 FC O2 FC O2 pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % pkat/mg % iPR 1749 75 11 134 597 128 16.6 105 430 86 105.7 291 2MeSiPR 3069 132 7.7 94 569 122 15.7 100 392 78 130.3 359 cZR 1614 69 - - 2.1 0.5 - - 246 49 - - 2MeScZR 2358 101 - - 4.5 1.0 - - 314 63 - - tZR 2716 116 48.2 588 149 32 4.4 28 193 39 46.2 127 2MeStZR 1979 85 42.2 515 433 93 1.6 10 293 59 59.3 163

Table 4: Substrate specificity of the vacuolar CKX enzymes and CKX5 of Arabidopsis in dehydrogenase (ferricyanide: FC) and oxygen (O2) mode at pH 6.0 given as absolute (pkat/mg) and relative activities. All data represent mean values of at least two biological replicates. Deviations between replicates did not exceed 10%; -: oxidative degradation of cis-derivatives can not be determined (see Materials and Methods).

Since R. fascians has a broad host range, we wondered if the same cytokinins accumulate in infected tissues of other plants. Therefore, we analyzed the cytokinin profiles of tobacco leafy galls and mock-inoculated tobacco plants. In contrast to Arabidopsis, no differential accumulation could be observed for cZ and 2MeScZ (see Figure 9). Classic cytokinin levels were up to 10-fold higher compared to 2-methylthio-type levels, and a more than 2-fold accumulation of iP was measured in leafy galls. Previously, also Eason et al. (1996) reported an accumulation of iP in infected pea tissue. Since iP is one of the bacterial cytokinins most abundantly produced in vitro, the higher levels observed in infected tobacco and pea tissues likely originate from R. fascians. Similar profiles were observed for the ribosides (data not shown). When the total amount of cytokinins (bases + ribosides) is calculated, a 62% increase is measured in infected tissues as compared to the controls.

Figure 9: Cytokinin profiles in mock inoculated control tobacco plants and leafy galls. Measurements were performed 4 weeks post infection; error bars represent SDs (n-3).

Identification and modus operandi of the bacterial cytokinins 53

The R. fascians cytokinins act synergistically.

The unusually strong outcome of infection with R. fascians, as compared to the relatively low cytokinin levels detected in supernatants and infected plant tissues, might be caused partly by a synergistic action of the cytokinin mix. Hence, the biological activity of the six identified cytokinin bases was tested as individual molecules and as equimolar mixes in different bioassays relevant in the context of the induced symptoms. First, we assayed the response of AHK4:GUS reporter lines to the R. fascians cytokinins. Plants were transferred (at developmental stage 1.05; see Materials and Methods) to medium containing 10 �M of the individual cytokinins or cytokinin mixes and histochemically analyzed after 10 days. iP, cZ, and tZ strongly activate the AHK4 expression, whereas their 2MeS- derivatives had a more moderate effect (see Figure 10). However, when plants were treated with an equimolar mix of iP, cZ, tZ and their 2MeS-counterparts (eq6) a stronger activation that equaled the levels obtained upon R. fascians infection was observed (see Figure 10).

Figure 10: Histochemical staining of AHK3:GUS and AHK4:GUS reporter lines upon treatment with 10 �M of the individual cytokinins or of the equimolar mixes of iP, 2MeSiP, cZ, and 2MeScZ (eq4), and of iP, 2MeSiP, cZ, 2MeScZ, tZ, and 2MeStZ (eq6).

Two other typical features of infected Arabidopsis plants are bleached leaves and anthocyanin accumulation (Depuydt et al., 2008). The contribution of the bacterial cytokinins to these phenotypes was determined by spectrophotometrical quantification of the chlorophyll and the anthocyanin content in the aerial parts of plants grown for 10 days on medium containing 10 �M of the cytokinins. Treatment with the individual cytokinins caused bleaching and anthocyanin accumulation to different degrees (see Table 5). Interestingly, when compared to

Chapter 3 54

the calculated additive effect of the individual cytokinins (add4 and add6), a synergistic effect was observed for eq4 (iP, cZ, and their 2MeS-derivatives), but especially for eq6 (see Table 5).

Treatment Chlorophyll (%) Anthocyanin (%)

control 100.00 ± 16.75 0.00 ± 0.00 iP 75.57 ± 5.84 38.34 ± 12.94 2MeSiP 77.45 ± 6.92 98.39 ± 4.63 cZ 86.28 ± 1.90 28.15 ± 3.81 2MeScZ 91.08 ± 4.16 48.93 ± 14.52 tZ 58.86 ± 6.59 38.87 ± 6.90 2MeStZ 86.83 ± 3.05 89.54 ± 14.22 Add 4 78.42 ± 14.21 53.59 ± 31.05 Add 6 80.10 ± 12.87 60.78 ± 31.33 Eq 4 71.85 ± 1.98 66.09 ± 22.92 Eq 6 51.85 ± 6.81 100.00 ± 10.32

Table 5: Loss of chlorophyll content and anthocyanin accumulation upon treatment with 10 �M of the individual cytokinins or of the equimolar mixes of iP, 2MeSiP, cZ, and 2MeScZ (eq4), and of iP, 2MeSiP, cZ, 2MeScZ, tZ, and 2MeStZ (eq6). The chlorophyll content and anthocyanin accumulation are represented relative to untreated (control) plants and upon eq6 treatment, respectively. add4, average number of events provoked by the four cytokinins at 1 �M as a measure for the additive effect of these cytokinins; eq4, number of events provoked by the equimolar mix at a final concentration of 10 �M (containing 2.5 �M of each CK). add6, represents the average number of events provoked by the six cytokinins at 0.1 �M as a measure for the additive effect of these cytokinins; eq6, number of events provoked by the equimolar mix at a final concentration of 1 �M (containing 0.167 �M of each cytokinin).

The most characteristic phenotype induced on all host plants upon R. fascians infection is the formation of shoots. To evaluate the shoot induction capacity of the bacterial cytokinins, rooting, green callus development, and shoot formation were scored at 28 days post incubation in a callus and shoot regeneration assay starting from Arabidopsis root explants (see Materials and Methods). In this assay tZ exhibited the highest shoot formation capacity, while 2MeScZ and cZ displayed the lowest activity (see Figure 11). Generally, the classical cytokinins exhibited higher activities than their 2-MeS counterparts, but interestingly, when compared to the calculated additive effect of the individual cytokinins (add4 and add6), for all three parameters, a synergistic effect was observed for eq4 at 10 �M (iP, cZ, and their 2MeS- derivatives) (see Figure 12A), and for eq6 at 1 �M (see Figure 12B). Identification and modus operandi of the bacterial cytokinins 55

Figure 11: Dose-dependent regeneration capacity of the individual and equimolar mixes of R. fascians-produced cytokinin bases. Root inhibition (triangle), and callus (square) and shoot induction (circle) events observed in the regeneration assays with classical (red symbols) and 2MeS (blue symbols) cytokinins. (A) iP and 2MeSiP; (B) cZ and 2MeScZ; (C) tZ and 2MeStZ; (D) eq4 (iP, 2MeSiP, cZ, and 2MeScZ; blue symbols) and eq6 (iP, 2MeSiP, cZ, 2MeScZ, tZ and 2MeStZ; red symbols).

Figure 12: Regeneration assay demonstrating the synergistic action of the R. fascians cytokinin bases. Equimolar mixes or additive effect of iP, 2MeSiP, cZ, and 2MeScZ (eq4, add4), and of iP, 2MeSiP, cZ, 2MeScZ, tZ, and 2MeStZ (eq6, add6). (A) Root inhibition, and callus and shoot induction events observed in the regeneration assays. add4, average number of events provoked by the four cytokinins at 1 �M as a measure for the additive effect of these cytokinins; eq4, number of events provoked by the equimolar mix at a final concentration of 10 �M (containing 2.5 �M of each cytokinin). (B) Root inhibition, and callus and shoot induction events observed in the regeneration assays. add6 represents the average number of events provoked by the six cytokinins at 0.1 �M as a measure for the additive effect of these cytokinins; eq6, number of events provoked by the equimolar mix at a final concentration of 1 �M (containing 0.167 �M of each cytokinin).

Chapter 3 56

This synergistic action of the R. fascians cytokinins was confirmed in a typical cytokinin bioassay: de-etiolation of dark-grown Arabidopsis seedlings. When Arabidopsis seedlings are grown in the dark, they show an etiolated phenotype. They have elongated hypocotyls, no chloroplasts develop and they are not able to form leafs. When relatively high amounts of cytokinins are added to the medium, de-etiolation occurs. The hypocotyls length is strongly reduced and they are able to form green leafs (Chory et al., 1994). After 4 weeks of dark treatment, the de-etiolation effect of the R. fascians cytokinins was quantified by scoring cotyledon opening and leaf formation, and measuring hypocotyl length. In this assay, the 2MeS-cytokinins exhibited lower activities than those observed for their classical counterparts. Importantly however, the number of leaf formation events was significantly higher for Eq6 at 10 �M and 1 �M than the calculated additive effect of the individual cytokinins (add6) (see Figure 13A and 13B). The hypocotyl length was divided into 3 categories, covering a range between 0 and more than 1.5 cm. Similarly, the number of inhibited hypocotyl events was significantly higher for Eq6 at 10 µM and 1 �M than the calculated additive effect of the individual cytokinins (add6) (see Figure 14A and 14B).

closed open leaf closed open leaf 10 0 % 10 0 %

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% iP tZ iP cZ tZ cZ Eq6 Eq6 Add6 Add6 control 2MeSIP 2MeStZ control 2MeSIP 2MeScZ 2MeStZ 2MeScZ

Figure 13: De-etiolation of Arabidopsis plants upon cytokinin treatment demonstrating the synergistic action of the R. fascians cytokinin bases. Equimolar mixes or additive effect of iP, 2MeSiP, cZ, 2MeScZ, tZ, and 2MeStZ (eq6, add6). The number of events (closed or open cotyledons, leaf formation) is given as percentages of the total number of germinated seedlings. (A) : Individual cytokinins; number of events scored at 10 �M; add6, sum of events scored for the six cytokinins at 10 �M; eq6, number of events scored for the equimolar mix at a final concentration of 60 �M (containing 10 �M of each cytokinin). (B) : Individual cytokinins; number of events scored at 1 �M; add6, sum of events scored for the six cytokinins at 1 �M; eq6, number of events scored for the equimolar mix at a final concentration of 6 �M (containing 1 �M of each cytokinin).

Identification and modus operandi of the bacterial cytokinins 57

� 0,5 cm � 1 cm >1 cm � 0,5 cm � 1 cm >1 cm 100% 100%

80% 80%

60% 60%

40% 40%

20% 20%

0% 0% iP iP tZ tZ cZ cZ Eq6 Eq6 Add6 Add6 control control 2MeSiP 2MeSiP 2MeStZ 2MeStZ 2MeScZ 2MeScZ Figure 14: De-etiolation of Arabidopsis plants upon cytokinin treatment demonstrating the synergistic action of the R. fascians cytokinin bases. Equimolar mixes or additive effect of iP, 2MeSiP, cZ, 2MeScZ, tZ, and 2MeStZ (eq6, add6). The number of events (hypocotyl length) is given as percentages of the total number of germinated seedlings. (A) : Individual cytokinins; number of events scored at 10 �M; add6, sum of events scored for the six cytokinins at 10 �M; eq6, number of events scored for the equimolar mix at a final concentration of 60 �M (containing 10 �M of each cytokinin). (B) : Individual cytokinins; number of events scored at 1 �M; add6, sum of events scored for the six cytokinins at 1 �M; eq6, number of events scored for the equimolar mix at a final concentration of 6 �M (containing 1 �M of each cytokinin).

Finally, in planta tissue proliferation is an essential aspect of the symptomatology (de O Manes et al., 2001), therefore the cell proliferative capacity of the six R. fascians cytokinins was quantified by using cytokinin-dependent tobacco callus and measuring the increase in fresh weight. The 2MeS-cytokinins proved to have a significant proliferative potential that was comparable to the classical cytokinins at 10 �M (see Figure 15). Interestingly, whereas a concentration of 100 �M of the classical cytokinins had a toxic effect on the callus cells, all three 2MeS-derivates continued to stimulate cell proliferation.

Figure 15: Dose response curve of the cytokinin- dependent tobacco callus growth. The growth of untreated (control) callus is represented by the dotted line. Error bars represent SDs (n=6).

Chapter 3 58

Discussion

Previously reported data did not allow to unequivocally conclude whether bacterial cytokinins were involved in R. fascians–induced symptom development (Goethals et al., 2001; Depuydt et al., 2008; Murai et al., 1980; Eason et al., 1996; Crespi et al., 1992; de O Manes et al., 2001; Galis et al., 2005). Here we addressed these questions by evaluating the response of Arabidopsis towards R. fascians infection at the level of cytokinin perception. Arabidopsis has three cytokinin receptors, AHK2, AHK3, and AHK4 with partly redundant but also divergent functions (Ueguchi et al., 2001; Higuchi et al., 2004; Riefler et al., 2006), and an AHK-receptor- independent pathway as evidenced by the intact, albeit reduced, plant body formation of an ahk2ahk3ahk4 mutant (Riefler et al., 2006). Functional analysis demonstrated that either AHK3 or AHK4 are required for symptom development, illustrating that both receptors are used in parallel for the recognition of the R. fascians cytokinins. Moreover, the non-responsiveness of the ahk3ahk4 mutant proved the crucial role of the bacterial cytokinins in the pathology. The AHK3 receptor is constitutively expressed in the shoot (Higuchi et al., 2004) and has the broadest ligand perception range (Spíchal et al., 2004); its high expression throughout the plant is not significantly modified by the bacteria. While AHK4 is thought to be part of a highly sensitive and specific recognition system (Spíchal et al., 2004) and AHK4 expression is mainly located in the root (Higuchi et al., 2004), we show that upon infection this gene is ectopically expressed in the shoot. Altogether, these data suggest that whereas AHK3 is likely the initial receptor for the R. fascians signals, AHK4 increases the sensitivity of the aerial plant parts for the bacterial morphogens. Like AHK3, AHK2 is highly expressed in shoot tissue (Ueguchi et al., 2001). However, the ahk3ahk4 double mutant proved to be much less responsive to cytokinin treatment than the other two double mutants, indicating that AHK2 has a lower sensitivity or controls a narrower downstream signaling cascade. These features might explain why this receptor has no crucial function in the R. fascians pathology. Already since 1966, researchers have wondered which bacterial cytokinins were responsible for the R. fascians–induced phenotype (Thimann and Sachs, 1966, Helgeson and Leonard, 1966). The different bacterial isolates had always been grown in relatively rich media, but from recent insights we know that essential virulence genes are not expressed under these conditions (Temmerman et al., 2000, Cornelis et al., 2002; Maes et al., 2001). Therefore, we determined the cytokinin profiles of the supernatants of the wild-type R. fascians strain D188 and its non-pathogenic derivative D188-5 as reference, grown in defined minimal medium optimized for virulence gene expression (Temmerman et al., 2000). Six cytokinin bases were detected: iP, 2MeSiP, cZ, 2MeScZ and, to a lesser extent, tZ and 2MeStZ. Riboside levels Identification and modus operandi of the bacterial cytokinins 59

were 10-fold lower but followed the same trends as the free bases. Given their low overall concentration and their lower biological activity (Spíchal et al., 2004), the ribosides are unlikely to contribute to the pathology. Besides 2MeStZ, all of these cytokinins had been detected in supernatants of different R. fascians isolates (Helgeson and Leonard, 1966; Scarbrough et al., 1973; Armstrong et al., 1976; Murai et al., 1980; Eason et al., 1996), but the use of near- isogenic strains differing only in the occurrence of the linear plasmid, allows a much stronger correlation between these cytokinins and symptom development. The observation that both strains produced the same cytokinin spectrum, likely explains the previous difficulties to attribute specific cytokinins to virulent strains (Murai et al., 1980; Eason et al., 1996). Nevertheless, the linear plasmid strongly contributes to the secreted levels of iP, 2MeScZ, and cZ. Currently, the consensus is that cZ mainly (Miyawaki et al., 2006) and 2MeS-derivates only (Murai et al., 1980; Prinsen et al., 1997) originate from tRNA degradation. As tRNA turnover rates are too slow (Klämbt et al., 1984) to account for the increased levels of iP, cZ, and 2MeScZ, we propose that these cytokinins are de novo synthesized by a linear plasmid- encoded machinery. The sequence of the linear plasmid of strain D188 did not reveal any cytokinin biosynthetic genes besides the fas operon (our unpublished results), and therefore, this is the only plausible locus to encode the cytokinin biosynthetic pathway (Crespi et al., 1992; Temmerman et al., 2000). The essential role of this pathway in the unique leafy gall pathology is further supported by the occurrence of a fas-like operon in the genome of the only other known leafy gall-inducing bacterium, S. turgidiscabies (Joshi and Loria, 2007). Unexpectedly, no significant differences were detected when bacteria were grown under non-inducing or inducing conditions for virulence gene expression. The inability of laboratory media to mimic the conditions present in plants is becoming increasingly evident, especially when intimate plant- pathogen interactions are considered (Marco et al., 2005) and is underlined by the observation that a different cytokinin spectrum is produced when pyruvate is added to the medium as a carbon source. If so, the measured cytokinin levels produced in vitro by R. fascians could be an underestimation of the true biosynthetic potential of the bacteria in planta. The accumulation of cZ and 2MeScZ in infected tissue, and the transient peak in the levels of iP, 2MeSiP, and 2MeStZ at the early phase of the interaction, might hint at biological significance of the identified bacterial cytokinins. The discrepancy with earlier reports in which the levels of cytokinins did not differ or even decreased upon infection (Depuydt et al., 2008; de O Manes et al., 2001), might be entirely attributed to the temporal and broader cytokinin profiling presented here, including the 2MeS-cytokinins. To understand the differential accumulation and decrease of the different cytokinins in infected tissues, we determined the substrate specificities of the three apoplastic CKX enzymes that function in cytokinin

Chapter 3 60

homeostasis. These enzymes were unable to efficiently degrade 2MeS-type cytokinins and cis- derivatives, probably contributing to the accumulation of cZ and 2MeScZ in infected Arabidopsis tissues. Interestingly, the vacuolar CKX enzymes and CKX5 (Werner et al., 2003) displayed a very different substrate specificity and degraded the 2MeS- and to some extent also the cis-derivatives much more efficiently. Possibly, these enzymes are involved in the removal of tRNA breakdown products during regular plant development. Intriguingly, when the cytokinin profile of tobacco leafy galls was determined, cZ and 2MeScZ did not accumulate, but iP was build-up instead, as similarly reported in R. fascians-infected pea plants (Eason et al., 1996). These observations suggest that the accumulation of specific bacterial cytokinins depends on the host, reflecting different substrate specificities of the plant homeostasis mechanisms. Consequently, by producing a spectrum of cytokinins, R. fascians may ensure that a subset of its morphogens is stable enough in planta to trigger plant developmental alterations. This strategy may well be at the basis of the very broad host range of R. fascians. Within the R. fascians-produced cytokinins, the biological activity of iP and tZ is generally recognized, whereas a true hormonal function has been reported for cZ in only a few plant species or organs (Spíchal et al., 2004, Yonekura-Sakakibara et al., 2004), and to date, none for 2MeS-type cytokinins (Matsubara, 1980). Here we show that the complete set of bacterial cytokinins are recognized by the cytokinin receptors AHK3 and AHK4 and have significant biological activities in five different bioassays, although the effect of the 2MeS-cytokinins was generally lower than that of their classical counterparts. More importantly, when plants were treated with equimolar mixes of the bacterial cytokinins clear synergistic effects were observed, for instance in shoot induction capacity. Since local cytokinin ratios are instrumental for regulating meristem activities (Zhao, 2008), in addition to the higher in planta stability, the production of cytokinin mixes could ensure a stronger impact on plant development. Moreover, in contrast to the classical cytokinins, the 2MeS-derivatives did not exhibit a cytotoxic effect within the tested concentration range in a tobacco callus bioassay. The continuous presence of the bacteria has been shown to be essential for sustaining symptoms (Vereecke et al., 2000) and a constant delivery of morphogens into the apoplast may require the need for the less active and less toxic 2MeS-derivatives. Based on the data presented, we propose the following model for the modus operandi of the R. fascians cytokinins in Arabidopsis. At the initial phase of the interaction, the bacterial cytokinins are perceived by AHK3 and in parallel, strongly activate AHK4 expression in the shoot, contributing to an increased sensitivity of the plant tissue. The synergistic action of the cytokinin mix leads to an intensified response, initiating the developmental alterations. To counter the induced changes, feedback and homeostatic mechanisms are activated in the plant, with the degradation of a subset of the bacterial cytokinins as a result. However, the Identification and modus operandi of the bacterial cytokinins 61

Arabidopsis CKX machinery is inadequate to eliminate all cytokinins of the bacterial mix and consequently, cZ and the less toxic 2MeScZ accumulate in infected tissues. These cytokinins are therefore responsible for the continuation of tissue proliferation and symptom maintenance throughout the interaction. In conclusion, our data have uncovered the enigma of the R. fascians pathology: the continuous challenge with a complex mixture of synergistically acting cytokinins eventually defeats nearly all plants and transforms them into shooty niches. Although other plant pathogens might use a similar strategy to manipulate the development of their hosts to their own advantage, to our knowledge, the “trick-with-the-cytokinin-mix” is a novel concept in phytopathology.

Materials and Methods

Bacterial strains and growth conditions.

The R. fascians strains used were the pathogenic D188 that carries the linear virulence plasmid pFiD188, and its nonpathogenic plasmid-free derivative D188-5 (Desomer et al., 1988). These strains were grown in liquid yeast extract broth (YEB) (Miller, 1972) at 28�C for 2 days under gentle agitation until late exponential phase. For the identification and quantification of cytokinins in bacterial supernatants, cultures were grown overnight under control and optimized conditions for virulence gene expression as previously described (Temmerman et al., 2000). A 2-days old R. fascians pre-culture was diluted 10 times in YEB medium. After overnight growth, the cells were washed and resuspended to OD600 2.0 in MinA medium (0.1% NH4SO4, 0.025%

MgSO4, 0.001% thiamine) adjusted to pH 5.0 by using citric acid and sodium citrate as a buffer system, supplemented with 20 mM succinate or pyruvate as carbon sources. To induce virulence gene expression 5 mM histidine was added.

Plant material and growth conditions.

A. thaliana and Nicotiana tabacum (L.) W38 seeds were sterilized by rinsing them for 2 min in 70% (v/v) ethanol, subsequently transferred to a 5% (w/v) NaOCl solution supplemented with 0.1 % (v/v) Tween20, and washed with sterile water. The seeds were germinated and grown on half (Arabidopsis) or full strength (tobacco) Murashige and Skoog medium in a growth

Chapter 3 62

chamber under a 16-h/8-h light/dark photoperiod at 21°C ± 2°C (Arabidopsis) or 24°C± 2°C (tobacco). The A. thaliana wild-type ecotype Columbia (Col-0) was obtained from the European Arabidopsis Stock Centre.

Infection, chemical treatment, and sampling.

Prior to infections 2 day old R. fascians cultures were washed and concentrated 4 times by resuspending them in sterile water. Infections and treatments of Arabidopsis were independently repeated 3 times on 50 plants at the developmental stage 1.05 (16 days old with five visible leaves) (Boyes et al., 2001). Plants were infected by applying locally a drop of bacterial culture to the shoot apical meristem and observed daily until 24 dpi for responsiveness scoring. For cytokinin analyses, the aerial plant parts were sampled by snap freezing in liquid nitrogen at 0, 2, 7, and 35 dpi. For chemical treatment, the cytokinins (OlChemIm Ltd., Olomouc, Czech Republic) were dissolved in dimethylsulfoxide (Sigma-Aldrich, St.Louis, MO) and applied at different concentrations ranging from 0.1 �M to 100 �M to Murashige and Skoog (MS) medium. Plants were transferred to this medium for 10 days and sampled for further analyses (phenotypic responsiveness, biological activity, and GUS expression). Infections of tobacco were biologically repeated 3 times. Three weeks old tobacco plants were decapitated by removing the apical meristem and the first axillary meristem and subsequently locally infected with bacterial culture or mock-inoculated with water. For cytokinin analyses, aerial plant parts (mock-inoculations) or leafy galls were collected 4 weeks post inoculation by snap freezing in liquid nitrogen.

Histochemical staining.

For GUS staining, the entire plant was sampled at 4 and 14 dpi or after 10 days of cytokinin treatment, and subsequently stained and analyzed as described in Depuydt et al. (2008). GUS-marked plants were submerged in 90% (v/v) acetone at 4°C for 1 h and transferred to a

GUS-staining solution of 2 mM 5-bromo-4-chloro-3-indolyl-ß-d-glucuronide and 0.5 mM K3

Fe(CN)6 in buffer containing 100 mM Tris and 50 mM NaCl (pH 7.0). After 19 h of incubation at 37°C in the dark, the tissue was cleared in 96% ethanol (v/v). Samples were stored in lactic acid (Acros Organics, New Jersey, USA) at 4°C until analysis with a binocular stereomicroscope (Zeiss, Jena, Germany). The three biological repeats each consisted of at least five plants per treatment.

Identification and modus operandi of the bacterial cytokinins 63

High Pressure Liquid Chromatography (HPLC) fractionation.

Cytokinins were extracted from plant material and purified according to Novák et al. (2003). Samples were fractionated using a reverse-phase column (4 x 250 mm, Vydac, 218 TP, C18, 300 Å, with guard column, Alltech) with water (A) and acetonitrile (B) as mobile phases. A linear gradient was set from 2% B to 50% B in 30 minutes and 50% to 100% B in 5 minutes, followed by a regeneration of 15 minutes in 2% B. The flow rate was 1ml/min and eluate was collected every minute. The collected fractions were dried in vacuo and tested for cytokinin activity in the yeast bio-assay described below.

Yeast bio-assay.

The yeast strain Saccharomyces cerevisiae sln1� carrying the plasmid p415Cyc1CRE1 was grown at 28°C on Yeast nitrogen base (YNB, 6,7g/L) medium supplemented with 2% galactose or and 0,65 g DO supplement -Ura/-His/-Leu and agar, to which 8mL 100mM His was added after autoclaving. Fifty µL of a water suspension containing S. cerevisiae sln1� [p415Cyc1CRE1] was diluted in 20 mL YNB medium supplemented with 2% glucose and 0,65 g DO supplement -Ura/-His/- Leu, to which 8mL 100mM His was added after autoclaving and 190 µL added to a 48-well plate. To test the fractions for their cytokinin activity, they were resuspended in 100 µL of 80% MeOH and 10 µL was added to each well. As a negative control water was added and as a positive control 10 µM tZ. The plates were incubated at 28°C and growth was scored after two days.

Bacterial and plant cytokinin measurements.

For measurement of classic cytokinin levels, cytokinins were isolated, purified and analyzed according to Novák et al. (2003) with some modifications. 2MeS-cytokinins were quantified by LC-MS in multiple reaction-monitoring mode. For cytokinin profiling of bacterial supernatants, 300 ml cultures were grown for 17 hours under control and optimized conditions for virulence gene expression as described above. Subsequently, 150 ml cell-free cultures were extracted by passing them over C18 cartridges, which were washed with water and eluted with C18 elution buffer (80% methanol, 2% acetic acid) and used for three technical replicates; these analyses were independently repeated 3 times. For the plant cytokinins, 1 g of plant material was used in three biological repeats. The occurrence of 25 cytokinin metabolites was evaluated, including isoprenoid and aromatic bases, their ribosides, and O- and N-glucosides.

Chapter 3 64

Escherichia coli receptor recognition assay.

E. coli KMI001 strains harboring plasmids pIN-III-AHK4 and pSTV28-AHK3 (Yamada et al., 2001, Suzuki et al., 2001) were used and assayed as described in Spíchal et al. (2004). E. coli pre-cultures, grown overnight to OD600 ~ 1.0 at 25°C in M9 medium enriched with 0,1% casamino acids (Sambrook et al., 1989), were diluted 50 times in 400 µl M9 medium containing 0,1% casamino acids. Cultures were incubated at 25°C with different cytokinin bases in concentrations from 0,1 to 100 µM and at the end of the incubation period 200 µl of the culture was centrifuged. One hundred µl supernatant was transferred to a test tube containing 2 µl of 50 mM 4-methylumbelliferyl galactoside and incubated at 37°C for 1h, after which the reaction was stopped by adding 100 µl 0,2M Na2CO3. Fluorescence was measured at excitation and emission wavelengths of 365 and 460 nm, respectively. The OD600 of the remaining culture was determined and the �-galactosidase activity was calculated as nmol 4-methylumbelliferone x -1 -1 OD600 x h .

Purification of recombinant CKXs.

Six AtCKX cDNAs were amplified by PCR without their signal sequences with the primers listed in Table 1 and cloned into the AvrII-linearized vector pGAPZ�::His(9) (Invitrogen) with the corresponding restriction enzymes (see Table 6). The constructs were subsequently stabilized into Pichia pastoris strain X33 by homologous recombination (Cregg et al., 1985) and, as such, the genes were constitutively expressed from the strong GAPDH promoter. The recombinant yeast strains were cultivated in YNB media buffered to pH 6.7 by potassium phosphate, supplemented with 2% glucose, at 28°C with orbital shaking at 230 rpm for 3 to 5 days. The cells were removed by centrifugation at 15,000 g and the pH of the cell-free medium was adjusted to pH 8.0 with 1 M Tris-base. The cell-free medium was concentrated to approximately 60 ml by ultrafiltration on a MiniKros Sampler Module (Spectrum, Rancho Dominguez, CA, USA) with a 10-kDa cut-off. Ultrafiltration was repeated three times to substitute the buffered media for 50 mM Tris/HCl, pH 8.0, supplemented with 20% ammonium sulfate. The concentrated proteins were loaded on an octyl-Sepharose CL-4B hydrophobic column (GE-Healthcare) connected to a BioLogic LP liquid chromatograph equipped with UV and conductivity detectors (Bio-Rad). After applying the sample, the column was washed with a decreasing step-gradient of ammonium sulfate and the eluate was fractionated. The fractions with enzymatic activity were pooled and concentrated with a stirred ultrafiltration cell (Millipore) equipped with a YM 10 membrane (10-kDa cut-off).

Identification and modus operandi of the bacterial cytokinins 65

Gene Primer sequence 5´ to 3´ end direction Restriction site

tccccgcggGTTCCAATCATTCTGTTAG SacII CKX1 acgcgtcgacTTATACAGTTCTAGG SalI

ggaattccatatgATTAAAATTGATTTACCTAAAT NdeI CKX2 gctctagaTCAAAAGATGTCTTGCCC XbaI

gtccatatgTCACACAACGAATTCGC NdeI CKX3 cggggtaccCTAACTCGAGTTTATTTTTTGA Asp718

tccccgcggCAGATGAGGGCATTGATG SacII CKX4 gctctagaTTAATTAAATATGTC XbaI

gtccatatgGTGGGTCTAAACGTG NdeI CKX5 ggggtaccTCACCATGAAGCCGC Asp718

gctctagaTCATGAGTATGAGACTGCCTTTTG XbaI CKX6 tccccgcggGCTTCTCTAGCAGCATTTC SacII

Table 6: Primer sequences used for PCR cloning of the AtCKX cDNAs into Pichia transformation vectors

The concentrated sample was applied onto a Bio-gel hydroxyapatite (Bio-Rad) column equilibrated with 10 mM potassium phosphate buffer (pH 7.7). The proteins were eluted by a linear gradient of 10 mM and 1 M potassium phosphate buffers (pH 7.7). Active fractions were concentrated to 2 mL and the buffer was replaced by 50 mM potassium phosphate (pH 7.4) containing 0.5 M NaCl with the ultra-filtration device. All CKX samples were finally purified on a Ni Sepharose HP (GE-Healthcare) equilibrated with 50 mM potassium phosphate buffer (pH 7.4) containing 0.5 M NaCl. His-tagged proteins were washed from the column by imidazole-containing (10 to 50 mM) loading buffer. Concentrated CKX proteins were checked for purity by SDS-PAGE followed by Western blotting with antibodies raised against barley CKX1 or CKX2 proteins. After purification, CKX proteins showed 20-80% homogeneity.

Cytokinin oxidase/dehydrogenase enzyme assay.

The CKX enzyme assay was based on the decolorization of the electron acceptor ferricyanide (500 �M) at 420 nm and normalized as other dehydrogenases (Appleby and

Chapter 3 66

Morton, 1959). The reaction mixture contained 100 mM McIlvaine buffer, pH 6.0, and 50 �M cytokinin substrate. The oxidase assay (Frébort et al., 2002) could not be done for the cis- derivatives, due to cyclization of the degradation product (Galuszka et al., 2007).

Chlorophyll and anthocyanin measurements.

Aerial plant parts were collected 10 days post cytokinin treatment, ground in liquid nitrogen, and 100 mg ground material used for per analysis. Chlorophyll was extracted and measured as described (Peng and Harberd, 1997). One mL 80% acetone was added to the grounded tissue, after which the tubes were shaken in the dark for 30 min. After centrifugation the optical densities of the supernatant were measured at 645 nm and 663 nm and chlorophyll content was calculated as chlorophyll (a+b)= 20,21*(A645) + 8,22*(A663). For anthocyanins, a slightly modified version of the method of Feinbaum and Ausubel (1988) was used. In short, total pigments were extracted for 10 min in 0.75 mL of 1% HCl/methanol, and 0.5 mL of distilled H2O was added. Chlorophyll was removed by chloroform extraction and the quantity of anthocyanin pigments in the aqueous/methanol phase was determined by measuring the absorbance at 530 nm minus that at 657 nm. All values were normalized to the fresh weight of each sample in three independent biological repeats.

Shoot induction assay.

The Arabidopsis plants were grown for 8 days, the roots harvested, cut in 1-cm explants and placed on 2.2 �M 2,4D for 6 days. Thereafter, they were placed on MS medium containing 0.9 �M IAA and different concentrations of the individual cytokinins or cytokinin mixes (ranging from 0.1 �M to 100 �M final concentration). As positive control, roots were placed on MS medium containing 0.9 �M IAA and 5 �M iP, and on 0.9 �M IAA as negative control. Phenotypes (root, callus, or shoot formation) were scored 28 days post inoculation and are represented as (events/root) x 100. When the entire root explant developed into callus or shoot, an arbitrary score of 10 was assigned. At least 20 root explants were used per treatment.

De-etiolation assay.

Ten Arabidopsis seeds were placed for 24 h in continuous light on MS medium containing 1 or 10 �M of the individual cytokinins or a mix of 1 or 10 �M of each of the six R. fascians cytokinins, and then transferred to the dark. In the presence of relatively high amounts of cytokinins however, the hypocotyl length is strongly reduced, the apical hook is lost, cotyledons Identification and modus operandi of the bacterial cytokinins 67

open, and green leaves are formed (Chory et al., 1997). After 4 weeks, cotyledon opening and leaf formation were scored. The assay was repeated 3 times and the observed phenotypes are represented as the percentage of events on the total number of germinated seedlings in the three experiments.

Tobacco callus bioassay.

The stimulation of cytokinin-dependent growth of tobacco callus (Nicotiana tabacum L. cv. Wisconsin 38) was measured as previously described (Holub et al., 1998), with slight modifications. The callus is maintained on low cytokinin concentration containing MS medium. For the assay the callus was placed in 6-well plates (3 ml of MS medium per well) on MS media containing 0,01 to 100 µM cytokinin for four weeks, after which it was weighed. The final concentration of DMSO in the media did not exceed 0.2%. Six replicates were prepared for each cytokinin concentration and the experiment was repeated at least twice.

Author contributions.

IP gathered data for Figures 11, 12, 13, 14 and Table 5, together with KV for Figures 3, 4, 5 and 9, and together with SD and KV for Figure 8. SD gathered data for Figures 1, 2, 7 and 10. LS gathered data for Figures 6 and 15. PG gathered data for Tables 1, 2, 3 and 4.

Chapter 4

�

Regulation and biochemistry of cytokinin biosynthesis in Rhodococcus fascians

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

“Biochemistry and biology of Rhodococcus� fascians cytokinin biosynthesis” � (in preparation). Pertry I, Václavíková K, Galuszka P, Spíchal L, Depuydt S, Temmerman W, Riefler M, Schmülling T, Strnad M, Holsters M, Tarkowski P, and Vereecke D.

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 71

Introduction

The link between cytokinin production and virulence of Rhodococcus fascians was the subject of a lengthy debate (Thimann and Sachs, 1966; Murai et al., 1980). More clarity on this issue came with the discovery of an ipt gene, located on the linear virulence plasmid pFiD188. A mutation of this gene in the D188-fas1 mutant resulted in a complete loss of pathogenicity (Crespi et al., 1992). The significance of the ipt in the pathology was further confirmed by the strict correlation of its presence with virulence in a high number of isolates tested. With the exception of one virulent strain, where the ipt gene resided on a large circular plasmid, in all other virulent strains the ipt gene was located on a linear plasmid (Stange et al., 1996). Despite the low overall sequence similarity (20 to 26% at protein level) with Ipts from Pseudomonas and Agrobacterium species, possibly reflecting different enzymatic specificities, Ipt activity has been demonstrated in extracts of Escherichia coli cells expressing the R. fascians ipt gene. Even more, the recombinant strains secreted significant amounts of cytokinins. Unexpectedly, no significant difference in cytokinin production was observed between the wild type R. fascians strain D188 and the D188-fas1 mutant. Subsequent expression analysis revealed that the ipt gene was only expressed under specific growth conditions, reflecting the presence of the plant. Indeed, when using these appropriate conditions Ipt activity was observed in D188 but not in fas1 (Crespi et al., 1992). The borders of the fas locus were determined by single homologous disruptive recombinations in the vicinity of the ipt gene, which identified a fragment of 6,5 kb of which sequence analysis revealed the presence of six open reading frames, fasA to fasF, with the same transcriptional orientation (see Figure 1). Although no transcription terminator could be recognized downstream of fasF, the next significant ORF is located 200 bp downstream of this gene. The six genes of the fas locus are thought to make up an operon because only small intergenic regions separate the fasA to fasD genes, and fasD-fasE and fasE-fasF are translationally coupled. A putative ribosome binding site, GAPuPuNGAPuTC, was identified immediately upstream of fasA, fasB, fasC and fasD (Crespi et al., 1994). A fasA insertion mutant (fas6) was not virulent although the mutation had no polar effect on the ipt expression, implying that the gene products of fasA, B or C or more likely their combinatorial action is required for full pathogenicity. FasA is similar to P450-type cytochrome mono-oxygenases and the amino-terminal region of FasB to 4Fe-3S type ferredoxins of Actinomycetes (Crespi et al., 1994). The carboxy-terminus of FasB on the other hand is homologous to the �-subunit of pyruvate dehydrogenase and FasC is similar to the �-subunit of this enzyme. Both proteins have a thiamine pyrophosphate binding site,

Chapter 4 72

which is required as a cofactor. The P450 cytochrome mono-oxygenase is hypothesized to hydroxylate the isopentenyl side chain of the cytokinin produced by FasD, while FasB and FasC would deliver the required energy via the ferredoxin-like domain by using pyruvate as an electron donor (Goethals et al., 2001). The fas5 mutant carries an insertion in the carboxy-terminal region of fasE and causes no, or less severe symptoms with outgrowing shoots on older plants, indicating that either fasE or fasF, or both, are required for full virulence (Crespi et al., 1994; Temmerman, 2000). Surprisingly, FasE showed homology to cytokinin oxidases (CKX) of A. thaliana. It was hypothesised that this protein was not involved in cytokinin degradation, but in biosynthesis by adding an additional modification to isopentenyladen(os)ine. The low sequence similarity of FasF with glutathione S-transferases would support this idea because it could point towards the introduction of glutathione or a different oligopeptide into the cytokinin molecule (Goethals et al., 2001). The importance of the fas operon in leafy gall induction, was underlined by its occurrence in Streptomyces turgidiscabies, a scab causing pathogen, were it was also linked with cytokinin production and leafy gall formation. The similarity between the corresponding Fas proteins is striking and the gene order is highly conserved (see Figure 1). Nevertheless, in S. turgidiscabies fasF was not part of the operon but separated by three additional genes encoding a transmembrane transporter and two putative methyltransferases (Joshi and Loria, 2007).

6 1 5

R. fascians D188

fasR fasA fasB fasC fasD fasE fasF

transmembrane transporter

mtr1 mtr2 fasF fasA fasB fasC fasD fasE

76 77 71 84 68 67 S. turgidiscabies

Figure 1: The fas locus consists of fasR, encoding an AraC-type transcriptional regulator, and the fas operon, which encodes the cytokinin biosynthetic machinery. The insertion mutants D188-fas1 (1) and D188-fas6 (6), and the fasR deletion mutant are non-virulent, while D188-fas5 (5) is only virulent on younger plants. The fas operon occurs also in S. turgidiscabies and is highly syntenic with that of R. fascians, although fasF is separated by a transmembrane transporter, and mtr1 and mtr2, which code for putative methyltransferases. The percent similarity of the S. turgidiscabies Fas proteins to those of R. fascians is indicated in the arrows representing the genes (modified from Joshi and Loria, 2007).

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 73

Expression of the fas genes is subjected to a complex regulatory network mediated by different regulatory loci that act at the transcriptional, but mainly at the translational level and translate multiple environmental factors, such as pH, carbon and nitrogen sources and cell density into appropriate production of the Fas protein machinery (Temmerman et al., 2000). Although transcriptional control is minor and transcription levels are not influenced by environmental conditions, based on expression data in different genomic backgrounds it appears that at least two transcriptional regulators controlling fas expression are present on the linear plasmid. The fasR gene is located approximately 2 kb upstream of the fas operon (see Figure 1) and encodes an AraC-type regulator that is essential for leafy gall formation and fas gene translation. The second regulator remains to be identified. Translation, on the other hand, is differentially controlled depending on the growth conditions. Under rich or control conditions, translational levels are generally low. Based on fas expression experiments under different conditions, a medium was defined, with pH 5.0 and a cell density of OD600 2.0, in which upon addition of gall extract or histidine combined with succinate or pyruvate the highest fas gene expression levels were obtained. This medium probably reflects the appropriate conditions met on the host. (Temmerman et al., 2000). Interestingly, although FasR is a transcriptional regulator, it proved to be essential for induction of fas gene translation suggesting that FasR indirectly regulates fas expression by controlling transcription of an unidentified translational regulator (Temmerman et al., 2000). Sequence analysis of the hyp locus of pFiD188 revealed a protein putatively involved in translational regulation and infection of plants with hyp mutants lead to the development of bigger leafy galls. Together, these findings suggest that the Hyp protein(s) may be the translational repressor(s) of the fas genes (Goethals et al., 2001). Finally, fas gene expression was also shown to depend on a functional att operon. Although it is currently unknown whether this is mediated by AttR, a LysR-type transcriptional regulator, or the autoregulatory compound synthesised by the att genes (Maes et al., 2001; Cornelis et al., 2002). Based on the available in vitro data, a model for fas gene expression was proposed (Temmerman et al., 2000). Under non-inducing conditions the fas genes are transcribed constitutively, but translation is blocked by a translational repressor (see Figure 2A). Under optimal inducing conditions the autoregulatory compound (AC) is synthesised by the gene products of the att operon. When a certain threshold is reached, a positive autoregulatory loop is activated via AttR so that more autoregulatory compound is formed (Maes et al., 2001), reaching a level required for binding and activating FasR. The FasR/AC complex will subsequently inhibit transcription of the translational repressor, relieving translation inhibition and allowing synthesis of the Fas proteins and the fas molecules (see Figure 2B).

Chapter 4 74

AttR

fas fasR fas att FasR - - translational translational + repressor repressor

A - B

Figure 2: Current model for fas gene expression. A. Under non-inducing conditions fas translation is repressed. B. Under inducing conditions, a positive autoregulatory loop activates att gene expression via AttR and production of AC ( : low concentration, high concentration) (indicated in black; Maes et al. 2001). This autoregulation is proposed to activate FasR, which would negatively control the transcription of a translational repressor of the fas genes (indicated in grey).

During the interaction with the host the fas genes are expressed throughout infection and mainly in the plant invading bacteria. The att genes on the other hand, are only expressed early in the interaction and in the epiphytic subpopulation. These data suggest that the autoregulatory compound produced by the Att proteins is only necessary for the initial expression of the fas genes (Cornelis et al., 2002) and implies that another, unknown regulatory mechanism controls the in planta expression of the fas genes. This complex regulation and strictly controlled fas gene expression reflects the importance of a timely production of the signal molecules. Possibly, cytokinin production imposes a significant energy demand on the bacterial metabolism and should only be activated in the presence of an appropriate host. In conclusion, it seems highly likely that the genes of the fas operon make up the cytokinin biosynthetic machinery. However, the role of each fas gene remains to be unravelled. In order to gain more insight in their putative contribution to cytokinin production, a detailed in silico and biochemical analysis was performed. Furthermore, we addressed the role of hyp, att and attR in fas gene expression in an attempt to complete the view on the fas gene regulation mechanism.

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 75

Results

In silico analysis of the fas locus reveals new homologies and genes likely involved in fas cytokinin biosynthesis

The fas locus consists of the transcriptional regulator fasR and the six genes of the fas operon and both are separated by 3282 base pairs (bp) (Crespi et al., 1994; Temmerman et al., 2000). With the program ORF Finder two open reading frames, ORF1 and ORF2, were identified in this locus, starting 611 bp downstream of fasR. The genes were separated by an intergenic domain of 40 bp and possibly make up a single transcriptional unit (see Figure 3), but the consensus sequence of the putative ribosome binding site was not detected upstream of their start codons.

Figure 3: Two open reading frames are located in between the fasR gene and the fas operon.

The R. fascians genome has an average G+C content of 61-68% (LeChevalier, 1986). It was previously reported that fasR has very low G+C content of only 53% (Temmerman et al., 2000). Interestingly, when we analysed the entire fas locus, all genes had a below average G+C content (see Table 1), which might point at a horizontal acquisition of the locus (see Chapter 6). Alternatively, the low G+C content might encompass a regulatory mechanism preventing efficient translation. To gain more or novel insights into the enzymatic functions of the Fas proteins and possibly the gene products of ORF1 and ORF2 in cytokinin production, similarity searches were performed with the BLASTP program against the non-redundant protein database at NCBI. An overview of the sequence characteristics and the best hits in the homology searches for each protein is given in Table 1. In case the same homologies were found compared to the data reported by Crespi et al. (1994) and Goethals et al. (2001), we show the amino acid sequence and indicate residues characteristic for the protein family. For newly identified homologies, amino acid alignments are given. Except for FasR, the best hits for all Fas proteins were those encoded by the fas operon of S. turgidiscabies (see Figure 1) (these hits are not indicated in Table 1).

Chapter 4 76

, 2007 , , 2006 , , 2007 , , 2007 , 2007 , 2007 , , 1998 , , 1997 , et al. et et al. et et al. et

et al. et al. et al. et 2001 al., et et al. et References References

product; %GC: GC GC %GC: product; Oliynyk Oliynyk Normand al. August et Oliynyk Morris, Powell and 1986 Werner Shirai Kurakawa Mongodin Mongodin

-20 -56 -61 -69 -90 - 20 -62 -42 -36

-116 4e 6e 1e 8e 3e 1e 4e 1e 3e 1e E-value

31/46 41/56 45/62 55/69 57/71 57/72 31/50 33/49 47/66 44/59

%ID/%Sim %ID/%Sim

Oryza sativa Oryza Pseudomonas Pseudomonas

Saccharopolyspora Saccharopolyspora Saccharopolyspora

KSM-K16 Homology

Arabidopsis thaliana Arabidopsis

locus. AraC family transcriptional regulator regulator transcriptional family AraC TC1 aurrescens Arthrobacter Methyltransferase erythraea Methyltransferase erythraea sp.CcI3 P450 Frankia Cytochrome subunit A transketolase Putative mediterranei Amycolatopsis subunit B Transketolase erythraea Saccharopolyspora Isopentenyltransferase syringae CKX1 decarboxylase Lysine (LOG) GUY LONELY (phosphoribohydrolase) Bacillus clausii clausii Bacillus

fas fas

59 52.9 56.5 58.6 59.6 60.4 55.7 58.4 60.5 %GC

bp/aa 834/277 834/277 852/283 852/283 915/304 939/312 768/255 597/198 1200/399 1200/399 1317/438

Stop 1397 2859 3750 5045 5973 6918 7714 9027 9623 (TAA) (TAA) (TGA) (TAG) (TGA) (TGA) (TGA) (TGA) (TGA)

(codon)

564 2008 2899 3846 5059 5980 6947 7711 9027 Start (ATG) (ATG) (ATG) (ATG) (ATG) (ATG) (ATG) (ATG) (GTG) (GTG) (codon) Sequence analysis and homologies of the of homologies and analysis Sequence

1 2 F A B E R C D fas mtr mtr fas fas fas fas fas fas

ORF Start: used start codon and position; Stop: used stop codon and position; bp/aa: nucleotide/amino acid length of the gene/gene gene/gene the of length acid nucleotide/amino bp/aa: and position; codon stop used Stop: position; and codon start used Start: content of the gene; %ID/%Sim:percentage identical/similar amino acids on overall protein sequence; E-value: expectation value. expectation E-value: sequence; protein overall on acids amino identical/similar %ID/%Sim:percentage gene; the of content 1: Table

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 77

The FasR protein is homologous to AraC-type transcription factors of different organisms (Temmerman et al., 2000; see Table 1), which act as positive or negative regulators and have reported regulatory functions in carbon metabolism, stress response and pathogenesis (Gallegos et al., 1997). These proteins consist of two structural domains, the first of which is a C-terminal conserved DNA-binding helix-turn-helix domain that contains a characteristic consensus sequence Ax5Sx3Lx3Fx4Gx10Rx3Ax3Lx8[I,V]x2[I,V]x4G[F,Y]x5Fx3F[R,K]x3Gx2P (x = any aa) conserved in FasR (see Figure 4). The N-terminus contains the effector/ multimerization domain, which is poorly conserved or even absent in certain members of this family and is involved in sensing environmental signals (Gallegos et al., 1997; Ibarra et al., 2008).

FasR

MTRLTARIVDGDSLTSIRCGPGEITRRQVDLEADPRFSVRIQYVLSGEVILSQNGVQISL KSGGIGLYYSDYSYTMTVTRASCLAVVTLRQRVGSREDFDAASGRGLRVVPNAEGTGKIL SDTLHSTSREIGMLNTFTAHNVISAVRLLAWEGMHPPHGHAPPDQQSRLVATARLMIEKN LSDPTLNPDLLARKLHVSVRQLHRAFEREERTLSAYIAYERIERCAADLIDPSLQQIPIG DISARWGLSDQSRLSRLFRELKGCSPSEFRRFYAEPS-

Figure 4: Amino acid sequence analysis of FasR. The HTH-motif is underlined and conserved residues of the consensus sequence are indicated in bold, non-conserved residues are shaded in grey.

ORF1 and ORF2 encode proteins with homology to S-adenosylmethionine-dependent methyltransferases (SAM- or AdoMet-MTase) of various organisms, such as Actinomycetes and plants, and were renamed mtr1 and mtr2 (see Table 1). Both Mtrs share 60% identity and 75% overall similarity (see Figure 6) and are most similar to and colinear with the two methyltransferases of S. turgidiscabies that are located immediately upstream of the fas operon of its pathogenicity island (see Figure 1), implying a possible role in fas molecule production and leafy gall formation. SAM-MTases catalyze the methylation of different substrates, such as hormones, lipids, proteins, and nucleic acids and can play a role in diverse biological processes, including signal transduction, detoxification, nucleic acid processing and biosynthesis. Using SAM as a methyl donor, different atoms, such as nitrogen, oxygen, carbon and sulphur are targeted for methylation. This protein family is divided into five different classes based on different structural folds and Mtr1 and Mtr2 belong to the largest Class I (Martin and McMillan, 2002; Schubert et al., 2003). Generally, MTases have limited sequence similarity. Although SAM binds at a similar position in the different proteins, the residues involved in SAM binding vary greatly, but two motifs are highly conserved - a glycine rich sequence E/DXGXGXG and an acidic loop - and a hydrophobic linker region, which is moderately conserved (Martin and McMillan, 2002). Some MT-ases

Chapter 4 78

contain a nitrogen-methylation-target-specific motif, [D/N/S]PP[Y/F], or a carbon-specific pattern, PC, (Schubert et al., 2003), but no target atom could be predicted for Mtr1 and Mtr2.

Se EG-DAATSHERTTRELHRLETWQAEFLLDHLGGVEPEHRIMDAGCGRGGSSFMAHERFGC 114 Sa DEPDPSRRHERITAELHRLEHAQAELLAGHLSPLSPADRVFDAGCGRGGGSVVAHLRYGC 119 Rf1 DS-VGEERENAILRELHRMENDQVGLILDALGPLPPNSRGMDAGSGRGGTSFRLAGATES 113 St1 DT-PPEHREEAILRELHRMESDQIRLILDALGTLPADSRGMDAGSGRGGTAFTIARELGH 113

Se SVEGVSLSRKQVDFANAQARERGVADKVAFHQLNMLDTGFDTASMRAIWNNESTMYVDLH 174 Sa HADGVTISAKQADFANEQARKRDIGDKVRYHHRNMLDTGLPTGAFAASWNNESTMYVELE 179 Rf1 RIDGVNFCEHHVAFAEQIARKRGWDTRVQFHLGNMLQAPFPDHTFDFVVSNETTMYADAY 173 St1 RVEGVNFCEHHVEFAEGLARKRGWDDRVRFHLGNMLQTPFEDGSFDFVVSNETTMYADAF 173

Se DLFAEHSRLLARGGRYVTITGCYNDVYGLPSRAVSTINAHYICDIHPRSGYFRAMAANRL 234 Sa LLFAEHARLLRRGGRCAVIIGCYNDTYGRASREVSLINAHYICDIHPRSAYFRAMARNRL 239 Rf1 EAMAEFSRLLRRGGRYVMTTWCRNDAVDPRSDATRQIDEHYVCRMHRRSTYFKAFAANGL 233 St1 EAMREFSRLLRPGGRYVMTTWCRDEAVDPRSEATRSIDKHYVCNMHRRGTYFQAFAAHGL 233

Se VPCAVVDLTEATVPYWRLRAKSP-LATGIEETFIEAYTSGSFQYLLIAADRV 285 Sa VPVHVEDLTGAALSYWELRKRSDRLATGIEDSFRTAYKNGSFQYLLIVADRV 291 Rf1 IPYRVAQYTHEAMPYWELRNNSK-LRTGVEDAFLSGYSDGSLNYLVIAAERI 283 St1 TPYRVERYTREAMPYWELRNLSA-LRTGVEEPFLEGYRDGSLNYMVVAAERI 284 A

Se -MTKSIHENGTA-----ASVYQGSIAEYWN-QEANPVNLELGEVDGYFHHHYGIGEPDWS 53 Sc MTTETTTATATAKIPAPATPYQEDIARYWN-NEARPVNLRLGDVDGLYHHHYGIGPVDRA 59 Rf2 MPNLDVAVLG------QHDVQQRRYWDAKKSDDINLLLGTEDGLYHHHYGIGDYDHS 51 St2 ---MDTVTLG------EHDAEQRGYWDAKQTDDINLLLGSEDGLYHHHYGIGDYDHA 48

Se VVEGDAATSHER-TTRELHRLETWQAEFLLDHLGGVEPEHRIMDAGCGRGGSSFMAHERF 112 Sc ALGDPEHSEYEKKVIAELHRLESAQAEFLMDHLGQAGPDDTLVDAGCGRGGSMVMAHRRF 119 Rf2 VLAASAELRESL-ILRELHRMESLEINLIVDALGEVSPSSRVMDAGSGRGGTTFTIADRF 110 St2 VLDAPAGRREEQ-IQRELHRMETLETGLVLDALGEVPSTARVMDAGSGRGGTSFVIADKF 107

Se GCSVEGVSLSRKQVDFANAQARERGVADKVAFHQLNMLDTGFDTASMRAIWNNESTMYVD 172 Sc GSRVEGVTLSAAQADFGNRRARELRIDDHVRSRVCNMLDTPFDKGAVTASWNNESTMYVD 179 Rf2 GCRVDGVNYCAHHVEFAEKLARERGSSDRVQFHFANMVQAPFEDNTFDYIVSNETTMCVD 170 St2 GCRVDGVNYCAHHVDFTQNLARERGVANRVQFHFTNMMQTPFEDGTFDFIVSNETTMCVD 167

Se LHDLFAEHSRLLARGGRYVTITGCYNDVYGLPSRAVSTINAHYICDIHPRSGYFRAMAAN 232 Sc LHDLFSEHSRFLKVGGRYVTITGCWNPRYGQPSKWVSQINAHFECNIHSRREYLRAMADN 239 Rf2 LGEAFTEFARLLRPGGRYVAVTWCRNDVVAERSEASRLIDEEYQCAMHTRSTYFQTLAAN 230 St2 IHKAFAEFSRLLRPGGRYVAVTWSRHDAVHPRSEASREIDEHYLCGMHKRSTYFEALAAN 227

Se RLVPCAVVDLTEATVPYWRLRAKSPLATGIEETFIEAYTSGSFQYLLIAADRV 285 Sc RLVPHTIVDLTPDTLPYWELRATSSLVTGIEKAFIESYRDGSFQYVLIAADRV 292 Rf2 GLVPYHVQRYTDEAIPYWDLRNQAALRTGVEEPFLQGFRERSIDYLVIACERL 283 St2 GLVPYHVRDHTEEAIPYWELREHSELRTGVEAPFLRGFREQSIDYLVIASERV 280 B

Figure 5: Sequence comparison of MTR1 (Rf1) (A) and MTR2 (Rf2) (B) with homologous proteins of Saccharopolyspora erythraea (Se; Accession YP_001105918), Streptomyces avermitilis (Sa; NP_822158), S. turgidiscabies (St1; AAW49309 and St2; AAW49310) and Streptomyces coelicolor (Sc; NP_631739). Numbers indicate amino acid positions. Identical amino acids ; similar amino acids ; grouped based on their structural features: GASTP (small and neutral), KRH, (polar and positively charged), WYF (aromatic), DENQ (acidic and polar) and VLIMC (non-polar). SAM-binding positions are underlined in black.

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 79

Mtr1 VMISADQYARDSYERELKAHWDAKTTDDINLLLGADDDLYHHHYAIGDFDRSILDSVGEER Mtr2 -MPNLDVAVLGQHDVQQRRYWDAKKSDDINLLLGTEDGLYHHHYGIGDYDHSVLAASAELR

Mtr1 ENAILRELHRMENDQVGLILDALGPLPPNSRGMDAGSGRGGTSFRLAGATESRIDGVNFC Mtr2 ESLILRELHRMESLEINLIVDALGEVSPSSRVMDAGSGRGGTTFTIADRFGCRVDGVNYC

Mtr1 EHHVAFAEQIARKRGWDTRVQFHLGNMLQAPFPDHTFDFVVSNETTMYADAYEAMAEFSR Mtr2 AHHVEFAEKLARERGSSDRVQFHFANMVQAPFEDNTFDYIVSNETTMCVDLGEAFTEFAR

Mtr1 LLRRGGRYVMTTWCRNDAVDPRSDATRQIDEHYVCRMHRRSTYFKAFAANGLIPYRVAQY Mtr2 LLRPGGRYVAVTWCRNDVVAERSEASRLIDEEYQCAMHTRSTYFQTLAANGLVPYHVQRY

Mtr1 THEAMPYWELRNNSKLRTGVEDAFLSGYSDGSLNYLVIAAERI Mtr2 TDEAIPYWDLRNQAALRTGVEEPFLQGFRERSIDYLVIACERL

Figure 6: Sequence comparison of Mtr1 and Mtr2 from R. fascians. Identical amino acids ; similar amino acids , grouped based on their structural features: GASTP (small and neutral); KRH, (polar and positively charged); WYF (aromatic); DENQ (acidic and polar) and VLIMC (non-polar). SAM-binding positions are underlined.

The FasA protein shows homology to cytochrome P450 enzymes (see Table 1, Crespi et al., 1994), which are heme-binding proteins involved in several oxidative reactions, such as C-hydroxylation, dealkylation and epoxide formation. P450 enzymes are involved in versatile cellular functions, including biosynthesis of steroids, antibiotics and defense compounds and are widespread throughout Archaea, Bacteria and Eukaryotes (Isin and Guengerich, 2007; Munro et al., 2007). The most characteristic P450 consensus sequence, FxxGxRxCxG, is located in the heme binding loop and the cysteine residue is absolutely conserved. [A,G]Gx[D,E]T[T,S] is another P450 consensus sequence and in between these two motifs, a third absolutely conserved motif ExxR is located (Werck-Reichhart and Feyereisen, 2000). All three consensus sequences are present in FasA (see Figure 7), although for the heme- binding motif the Arg residue is replaced by the similar His residue. The enzymatic activity of most P450 cytochromes is dependent on electrons generated by different types of redox partners and delivered via ferredoxins or flavodoxins (McLean et al., 2005). Most likely, electrons for FasA are delivered via the N-terminal ferredoxin-related region of FasB, which is a dual protein of which the C-terminal region is homologous to the transketolase subunit A. FasC is homologous to the B subunit of this enzyme (see Table 1 and Figure 7) (Crespi et al., 1994). Transketolases transfer a two-carbon unit from a ketose to an aldose, link the pentose pathway and glycolysis, provide precursors for nucleotide, aromatic amino acid and vitamin biosynthesis, and play an important role in the Krebs cycle in plants (Kochetov and Sevostyanova, 2005). Although P450 systems in prokaryotes are generally soluble (Werck-Reichhart and Feyereisen, 2000), the predicted presence of a

Chapter 4 80

transmembrane domain in FasC suggests that in R. fascians the FasA-B-C system might occur associated with the membrane.

FasA MAGTADLPLEMRRNGLNPTEELAQVRDRDGVIPVGELYGAPAFLVCRYEDVRRIFADSNR FSNAHTPMFAIPSGGDVIEDELAAMRAGNLIGLDPPDHTRLRHILAAEFSVHRLSRLQPR IAEIVDSALDGLEQAGQPADLMDRYALPVSLLVLCELLGVPYADRDELRDRTARLLDLSA SAEQRAVAQREDRRYMATLVTRAQEQPGDDLLGILARKIGDNLSTDELISIISLIMLGGH ETTASMIGLSVLALLHHPEQAAMMIEDPNCVNSGIEELLRWLSVAHSQPPRMAVTEVQIA GVTIPAGSFVIPSLLAANRDSNLTDRPDDLDITRGVAGHLAFGHGVHFCLGHSLARMTLR TAVPAVLRRFPDLALSPSHDVRLRSASIVLGLEELQLTW-

FasB MKVVVNERRCFGSGQCVLVAPEVFEQSNDGTVTLLVDKPSPDNHSLVRAAARSCPATAIR FEENAMRQEPTEFSYDDLPALISRMRGDERHSFSSSSTMDVLWVLYDEIPNVSPESPDDD DRDRFLLSKGHGPMAYYAVLAAKGFLRPELLDTWATKNSPLGFAPDRTKISGVEMSGGSL GHGLPLAVGVAMGLRIQNRHAPRVFVLIGDGEFDEGSNHEAMAFAGRARLNQLTVIVLDN GTASMGWPHGIDKRFDGEGWDTININGADHEEIAAALNRDHNDRPLAVVATVTRQSARSS IQQR-

FasC

MNSADTQEPKSFNHTDMWTAFGTTMSGALETDPRAVVVLADIGAHLFKAAAIADPNRVIN VGIREQLMMGVAGGLAMCGMRPVVHTVAAFLVERPLEQIKLNFAQQDVGAVLVSWGASYD LSEFAFSHFTPGDITVIDSMPNWTVHVPGHPQEAADLLLESLPGDGRVYLRLSSQVNRYP HAVRGTSFTPIKYGTRGVVLAVGPCLDAVLSATSMLDVTILYAATIRPFDATGLCAAVQA VNRPNVVLVEPYLAGTSAHQVSSSLVSHPHRLLSLGVRREMEDRHYGTPDDHDHIHGLDA RSLSNSINSFLG-

Figure 7: Amino acid sequence analysis of the fas encoded proteins. FasA shows homology to cytochrome P450s. The consensus sequences are underlined and conserved residues are indicated in bold, the conserved cysteine residue is shaded in grey. FasB contains a N-terminal ferredoxin domain, indicated in bold, and a carboxyterminal transketolase domain, which is underlined. FasC shows homology to a transketolase domain, which is underlined.

FasD is a functional Ipt (Crespi et al., 1992). Based on the amino acid sequences of several bacterial Ipts and tRNA-Ipts a common Ipt consensus pattern was described:

GxTx2GK[S,T]x5[V,L,I]x7[V,L,I][V,L,I]x2Dx2Qx57-60[V,L,I][V,LI]xGG[S,T] (x = any aa) (Kakimoto, 2001), which was also found in FasD (see Figure 8). The FasE protein shows homology to several plant cytokinin oxidases (Temmerman, 2000; Goethals et al., 2001; Table 1), which irreversibly degrade cytokinins. FAD is used as a cofactor and binds at an N-terminal domain with a conserved GHS-motif (Schmülling et al., 2003), present in FasE (see Figure 8). A specific residue present in the cytokinin binding domain is hypothesised to control substrate specificity by interacting with the N9 atom from the purine ring. When the enzyme is specific for cytokinin bases, usually Glu is present at Regulation and biochemistry of cytokinin biosynthesis in R. fascians 81 this position, while Ala or Ser indicate a preference for N9-glucosides or cytokinin nucleotides (Galuszka et al., 2007; Frébort, personal communication). FasE however, has a Gln at this position, which is similar in size to Glu but bears an opposite charge that would repulse the N9 atom. This might reflect a different substrate specificity for FasE.

FasD MKESTMAQTQARFDRVRWEPGVYAIVGATGIGKSAEASKLALSHSAPIVVADRIQCYSDL LVTSGRAFDAKVEGLNRVWLDNRTIHQGNFDPDEAFDRLIKVLTSYVDRGEAVVMEGGSI SLILRFAQTISNLPFPAVVNVMPIPDRQHYFAQQCARARQMLRGDSTGRNLLTELAEAWV LGDQHNFIASVAGLDCVLDWCATHSVTPEELANRDLTTEVLDELAASMGGRYVEHGVLQQ EIFLRTFGAPGVTAR-

FasE MSGIWHTDDVHLTSAGADFGNCIHAKPPVVVVPRTVADVQEALRYTAARNLSLAVRGSGH STYGQCQADGGVVLDMKRFNTVHDVRSGQATIDAGVRWSDVVAATLSRQQTPPVLTDYLG TTVGGTLSVGGFGGSSHGFGLQTDNVDSLAVVTGSGDFRECSAVSNSELFDAVRGGLGQF GVIVNATIRLTAAHESVRQYKLQYSNLGVFLGDQLRAMSNRLFDHVQGRIRVDADGHLRY RLDLAKYFTPPRRPDDDALLSSLQYDSCAEYNSDVDYGDFINRMADQELDLRHTGEWFYP HPWASLLIPADKIEQFIETTSSSLTDDLGNSGLIMVYPIPTTPITAPFIPIPHCDTFFML AVLRTASPGAEARMIASNRLLYEQARDVGGVAYAVNAVPMSPGDWCTHFGSRWQAIARAK RRFDPYRILAPGYRMSFD-

Figure 8: Amino acid sequence analysis of the fas encoded proteins. FasD encodes an isopentenyltransferase. The predicted consensus motif is underlined and conserved residues are indicated in bold. FasE shows homology to cytokinin oxidases and contains the typical FAD-binding domain, indicated in bold, with the conserved GHS-motif and a cytokinin binding domain, which is underlined. The residue, which is hypothesised to control substrate specificity, is indicated.

Previously, the FasF protein was reported to be homologous to glutathione S- transferases and was hypothesised to be involved in the transfer of glutathione or another peptide side chain to the adenine moiety (Goethals et al., 2001). The recent similarity searches revealed that FasF strongly resembles lysine decarboxylases (LDC) from different organisms (see Table 1, Figure 9). Members of this family share a highly conserved motif

PGGxGTx2E (Jeon et al., 2006), which is also found in FasF (see Figure 9). LDCs are involved in the production of the polyamine cadaverine (Phan et al., 1982). Polyamines are known to influence plant morphogenesis and regeneration capacity (Shoeb et al., 2001). Possibly, FasF might be responsible for the biosynthesis of a polyamine, which in combination with the fas cytokinins could regulate leafy gall formation. When we analysed supernatant of bacterial cultures grown under conditions optimal for virulence gene expression, no cadaverine was detected. Moreover, the very low levels of cadaverine in control Arabidopsis and tobacco plants did not significantly differ upon infection (data not shown). Together, these findings argue against an LDC activity of FasF.

Chapter 4 82

Interestingly, FasF is also homologous to the LONELY GUY (LOG) protein of rice (59% overall similarity, see Table 1 and Figure 9), which exhibits phosphoribohydrolase activity towards cytokinin nucleotides, resulting in the direct release of free bases (Kurakawa et al., 2007). This homology indicates that FasF possibly mediates cytokinin production in R. fascians by the release of free cytokinin bases from their nucleotide precursors.

Rf ---MNLRPMPATTVS------AQARPTPKSVTVFCGAMPGRGTKYGQLAEGM 43 St ------MKCVTVFCGATPGLAGIHTRAAAEL 25 Ar ------Os -MAMEAAAERSAGAGAAATAAPESGGGGAGERRSRFRRICVYCGSAKGRKASYQDAAVEL 59 At ---MEIKGE------SMQKSKFRRICVFCGSSQGKKSSYQDAAVDL 37

Rf GRAIARSKLRLVYGGARVGLMGTLANAALDSGGTVVGVIPESFTAIPEAAHHG--LTELH 101 St GSAIARNGLRLVYGGATVGLMGTLADAALGAGGVVVGVIPQYLSVVPEAAHPG--LTELH 83 Ar ---MARSGIGLVYGGASIGLMGAIADAARSDGGEVIGVIPRALAEK-EIAHTD--LADLR 54 Os GKELVERGIDLVYGGGSIGLMGLVSHAVHDGGRHVIGVIPKSLMPR-EVTGEP--VGEVR 116 At GNELVSRNIDLVYGGGSIGLMGLVSQAVHDGGRHVIGIIPKTLMPR-ELTGET--VGEVR 94

Rf VVHDMHQRKALMAELGDAFIALPGGVGTAEEFFEVLTWSHLGLHNKPCVLLNDNEYYRPL 161 St VVPDMHRRKAMMAELGDAFVALPGGVGTAEEFFEVLTWSHLRLHDKPCVLLDIHGYYRPL 143 Ar VVETMHERKALMAALSDGFIALPGGLGTLEELFEVWTWAQLGYHNKPCALLDIAGFYKRL 114 Os AVSGMHERKAEMARFADAFIALPGGYGTLEELLEVITWAQLGIHKKPVGLLNVDGFYDPF 176 At AVADMHQRKAEMAKHSDAFIALPGGYGTLEELLEVITWAQLGIHDKPVGLLNVDGYYNSL 154

Rf LSYIEHAAVEGFITPATRSRVIVCKDIEGAIAAIR--SP------198 St LAFIEHAVREGFIDSDTARRVIVCERVEQVFEVLR--HVGIPGG------IDTADA---- 191 Ar DSFLDHVVGEAFLTASHRNILLVEEDAEVLISAMANDSATLKSR------SEESDFRRLR 168 Os LSFIDMAVSEGFIAEDARRIIISAPTARELVLKLEEYVPEYEVGLVWDDQMPHSFAPDLE 236 At LSFIDKAVEEGFISPTAREIIVSAPTAKELVKKLEEYAPCHER------VATKLCWEME 207

Rf ------St ------Ar R------169 Os TRITSS-- 242 At RIGYSSEE 215

Figure 9: Alignment of FasF from R. fascians (Rf) with Fas6 from S. turgidiscabies (St; AAW49312), Riorf52 from A. rhizogenes (Ar; NC_002575.1), LOG from Oryza sativa (Os; AK071695) and a LOG homologue At2g37210 from A. thaliana (At; NM_129277). Identical amino acids ; similar amino acids , grouped based on their structural features: GASTP (small and neutral), KRH, (polar and positively charged), WYF (aromatic), DENQ (acidic and polar) and VLIMC (non-polar). Numbers indicate amino acid positions. The lysine decarboxylase motif PGGxGTX2E is boxed in black. The sequence used for the rice LOG phylogenetic analysis (Kurakawa et al., 2007) is underlined in black dots.

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 83

Biochemistry of the Fas proteins

The FasD isopentenyltransferase produces iP and zeatin-type cytokinins in vitro

Isopentenyltransferase activity has been demonstrated in crude protein extracts of E. coli cells expressing fasD and of induced R. fascians D188 cultures, using AMP as a side chain acceptor and DMAPP as a side chain donor. No Ipt activity could be measured in protein extracts of the D188-fas1 mutant (Crespi et al., 1992). Here, we did a more thorough biochemical characterisation of FasD, by heterologously expressing and purifying an N- terminal His-tagged version of the protein and testing different side chain donors and acceptors and different reaction conditions (see Materials and Methods). To determine substrate specificity, the Km values of different donor and acceptor substrates were assayed at 25°C and pH 8.0, standard conditions for IPT measurements (Takei et al., 2001; Sakano et al., 2004). Because the Ipts from A. tumefaciens can produce both iP and tZ by using, respectively, DMAPP or HMBPP as side chain donors and AMP is side chain acceptor (Krall et al., 2002), we tested both donor compounds in combination with AMP, ADP or ATP as acceptors. The data presented in Table 2 show that purified FasD produced iPR in vitro, confirming the results obtained by Crespi et al. (1992). Generally, bacterial Ipts use AMP as preferred side chain acceptor (Kakimoto, 2003), but interestingly FasD most efficiently prenylated ADP, and therefore resembles more plant IPTs (see Table 2). Almost no tZ could be detected in supernatant of induced D188 cultures (see Chapter 3), nevertheless, unexpectedly FasD very efficiently produced tZ in vitro with all three acceptors, albeit with a slight preference for AMP, thus resembling bacterial Ipts (see Table 2).

Acceptor AMP ADP ATP Donor

DMAPP 210 (± 0.067) 70 (± 0.02) 290 (± 0.08)

HMBPP 4 (± 0.001) 10 (± 0.003) 10 (± 0.005)

Table 2: Km values (nM) of FasD for different side chain acceptors with DMAPP or HMBPP as side chain donors. Reactions were performed at 25°C, pH 8.0 for 1h. Errors represent SD (n=3).

We next tested the pH optimum for FasD by measuring iPR production at 25°C, using DMAPP as side chain donor and AMP as side chain acceptor. pHs ranging from pH 3.5 to 7.5 were obtained in citric acid phosphate buffer and from 7.5 to 9.0 in Tris buffer. The results in Figure 10 show that the pH optimum of FasD lies between 5.5 and 7.5 with a peak at pH 5.5.

Chapter 4 84

Citric acid-phosphate buffer Tris buffer 180 160 140 120 of protein of

-1 100 Figure 10: Effect of pH on FasD mg 80 -1 IPT activity. Reactions were 60 performed at 25°C, for 1h with 40 AMP and DMAPP as substrates, and citric acid phosphate or Tris µM iPR h iPR µM 20 solution as buffer. 0 Error bars represent SD (n=3) 3456789 pH

The fasE gene encodes a functional cytokinin dehydrogenase

Although fasE shows homology to CKXs of different plants, only circumstantial proof was available for its activity. Supernatant of D188, D188-fas1 and D188 cultures overexpressing fasA to fasF incubated with radioactively labeled adenine subjected to thin layer chromatography (TLC) analysis, revealed the presence of a fas-dependent spot. When a similar experiment was performed with D188 overexpressing only fasA to fasD the intensity of this spot increased, suggesting that the FasE protein degrades fas-derived products (Den Herder and Temmerman, 2002; unpublished results). Moreover, addition of increasing amounts of thidiazuron, known to inhibit CKX activity (Chatfield and Armstrong, 1986), during the feeding experiment, resulted in a gradual increase in the intensity of the fas-dependent spot. These data suggest that fasE encodes an active CKX that controls the level(s) of fas- derived cytokinin(s) (Temmerman, unpublished results). In the absence of a suitable electron acceptor, CKX proteins can use oxygen and function as oxidases, although under these conditions the reaction rates are lower (Galuszka et al., 2007). To directly test if FasE is a functional CKX, a plasmid, pMALfasE, was constructed expressing a fusion protein between the maltose binding protein (MPB) and FasE for heterologous expression in E. coli (Den Herder and Temmerman, 2002). Purified FasE-MBP was treated with Factor Xa, cleaving the MBP tag and releasing FasE from the fusion protein. The enzyme activity of both protein preparations was tested at the physiological pH 6.0, in oxidase mode, using oxygen and in dehydrogenase mode, using ferricyanide (FC) as the electron acceptor. The capacity of the R. fascians CKX enzyme to degrade different cytokinin substrates was expressed as absolute enzymatic activities or as relative values compared to their ability to degrade iP. No significant differences were observed for FasE- Regulation and biochemistry of cytokinin biosynthesis in R. fascians 85

MBP and FasE, indicating that the MPB tag does not influence the activity (data not shown). The data in Table 3 illustrate that FasE preferentially acted as a dehydrogenase and efficiently degraded iP, iPR, iPRMP and 2MeSiP, but not iP-9-glucoside (iP9G), tZ, 2MeStZ, cZ or 2MeScZ.

ferricyanide O2

pkat/mg % pkat/mg % iP 183.1 100 34.8 100 iPR 162.9 89 36.5 105 iPRMP 246.8 135 54.2 156 iP9N 5.1 3 8.1 23 2MeSiP 299.9 164 37.2 107 cZ 45.2 25 - - 2MeScZ 30.4 17 - - tZ 18.9 10 3.4 10 2MeStZ 14.3 8 5.1 15

Table 3: Substrate specificity of the FasE in dehydrogenase (500 µM FC) and oxygen (O2) mode at pH 6.0 given as absolute (pkat/mg) and relative activities. All data represent mean values of four biological replicates. Deviations between replicates did not exceed 15%. -: oxidative degradation of cis-derivatives can not be determined (see Materials and Methods).

We next determined the effect of different electron acceptors at varying concentrations on the reaction rate, which was expressed relative to the basal reaction rate of FasE in oxidase mode. These experiments confirmed that FasE is an efficient dehydrogenase and show that besides ferricyanide, 2,3-dimethoxy-5-methyl-1,4-benzoquinone (Q0), and to a lesser extent 2,6-dichlorophenolindophenol (DCIP) are suitable electron acceptors (see Table 4), indicating that the natural electron acceptor most likely will be a quinone, resembling plant CKXs (Frébortova et al., 2004).

FasE

oxygen 1x 500 µM DCIP 1.8x 50 µM DCIP 2.7x 500 µM Q0 10.5x 50 µM Q0 6.2x 500 µM FC 5.4x 50 µM FC 1.5x

Table 4: Relative reaction rates of FasE in dehydrogenase mode (200 µM iPR, pH 6.0).

Chapter 4 86

When R. fascians D188 was grown under inducing conditions only 0.01 pkat/mg CKX activity could be measured in crude protein extracts (reaction performed with FC and iPR). Because the relative concentration of FasE in the total protein extract is not known, this number might be an underestimation of the specific activity of FasE in R. fascians cells.

FasF is possibly a phosphoribohydrolase

To test if FasF is a functional lysine decarboxylase, a plasmid, pMALfasF, was constructed expressing a fusion protein between the maltose binding protein (MPB) and FasF for heterologous expression in E. coli. Purified FasF-MBP fusion protein displayed only a low LDC activity (2.16 ± 0.42 nmol cadaverine mg-1hour-1) compared to a positive control sample (11.65 ± 4.22 nmol cadaverine mg-1hour-1) (Den Herder and Temmerman, 2002), which indicates that FasF unlikely functions as a lysine decarboxylase, confirming the data obtained by profiling the polyamine levels in induced D188 and infected plants. FasF is also homologous to a phosphoribohydrolase from rice. Preliminary data suggest that purified MBP-tagged FasF protein is capable of releasing iPMP. However, the efficiency is rather low which might be caused by the presence of the fused MBP-tag or a different substrate specificity (data not shown). In order to draw solid conclusions the assay will have to be optimised.

Novel insights in fas gene regulation

The fas genes are differentially expressed in vitro

Based on the expression patterns of fasA and fasD, fas gene translation is apparently tightly controlled (Crespi et al., 1992; Temmerman et al., 2000). no induction could be observed in cytokinin production when using these optimized conditions (Chapter 3). To get a more complete view on the expression of the different functional units encoded by the fas locus, non-replicating constructs were generated carrying translational uidA fusions with mtr1, mtr2, fasA, fasD, fasE and fasF. These constructs were introduced into R. fascians D188 via electroporation and in the resulting clones, the correct single homologous recombination was verified by Southern hybridisation. The different clones were grown under inducing (histidine/succinate) and non-inducing conditions (succinate) for 25 hours and GUS activity was measured. Although some expression could be measured for fasD under non-inducing conditions, the levels reached upon induction were 4-fold higher, which is reflected by a small but Regulation and biochemistry of cytokinin biosynthesis in R. fascians 87 significant induction in iP production (see Chapter 3). Similarly, fasA expression was strongly inducible, albeit that the maximum level was more than 6-fold lower than that of fasD (see Figure 11). These data are largely in agreement with the data reported by (Crespi et al., 1992; Temmerman et al., 2000). Unexpectedly however, under the conditions tested, hardly any expression could be measured for mtr1, mtr2 and fasE, and although fasF expression was 7-fold induced, the level was very low (see Figure 11). Possibly expression of mtr1, mtr2, fasE and fasF is triggered by other conditions that reflect another stage during the interaction with the plant. Altogether, these data show that the fas genes are differentially expressed in vitro adding another level of complexity to fas gene regulation and in vitro cytokinin production.

20 NI I 18

16

14

12

10

8

GUS GUS activity (u) 6

4

2

0 mtr1 mtr2 fasA fasD fasE fasF

Figure 11: Inducibility of expression of different integrated translational fas-uidA fusions under non- inducing (succinate) (NI) and inducing (histidine/succinate) (I) conditions. mtr1: D188:mtr1, mtr2: D188:mtr2, fasA: D188:fasA, fasD: D188:fasD, fasE: D188:fasE, fasF: D188:fasF. Error bars represent SD (n=2).

The att operon plays a dual role in fas gene expression

Based on the working model for fas regulation, fas gene expression requires a high concentration of autoregulatory compound produced by the Att proteins, which is obtained by a positive feed-forward loop activated after initial expression of the att operon (see Figure 2). To test this hypothesis, the induction level of fasA translation was determined in the wild type strain D188 and its plasmid-free derivative D188-5 as controls, in D188-�att, which carries a deletion of the nine att genes, and in D188-�attR, in which the attR gene is deleted. In both latter strains, the autoinduction of the att operon is lost (Maes et al., 2001). The construct

Chapter 4 88

pJDGV5, a replicating vector carrying a translational fasA-uidA fusion (Temmerman et al., 2000), was introduced via electroporation into these strains, the resulting clones were grown under two inducing (gall extract/succinate or histidine/succinate) and two non-inducing conditions (succinate and plant extract/succinate), and GUS activity measured (see Materials and Methods). Typically, the expression levels under inducing conditions were highly variable, nevertheless the inducibility of the fasA gene was completely lost in the D188-�att and D188-�attR background (see Figure 12), indicating that a functional att operon is required for induction of fas gene translation.

45 succ plant+succ 40 gall+succ his+succ

35

30

25

20

15 GUS activity (u) activity GUS 10

5

0 D188 D188-5 D188-�att D188-�attR Figure 12: Inducibility of translation of a fasA-uidA fusion on pJDGV5 in different genomic backgrounds grown under non-inducing (succ: succinate, plant+succ: plant extract/succinate) and inducing (gall+succ: gall extract/succinate, his+succ: histidine/succinate) conditions. Error bars represent SE (n=10).

To determine whether this expression pattern was the result of the loss of the feedback loop or of a required transcriptional regulation by AttR or a combination of both, a replicating plasmid pIPRF1 was generated that carried the attR gene with its own expression signals and the same translational fasA-uidA fusion present in pJDGV5. This construct was introduced via electroporation into the four test strains and expression monitored under inducing and non-inducing conditions. Again fas gene expression was highly variable under inducing conditions and in strain D188-5 no induction occurred (see Figure 13). Interestingly, in strain D188 the induction levels were much higher than those obtained for D188[pJDGV5], suggesting a direct involvement of the AttR protein on fas gene expression. Indeed, introduction of pIPRF1 into D188-�attR restored inducibility of fasA to a level that was 2-fold higher than in D188[pJDGV5] (see Figure 13). These data imply that AttR may be the second, unidentified transcriptional regulator controlling fas gene transcription, besides Regulation and biochemistry of cytokinin biosynthesis in R. fascians 89

FasR. Since AttR is a LysR-type transcriptional regulator and these proteins act by binding a specific recognition sequence, T-N11-A, surrounded by short inverted repeats (Goethals et al., 1992), we searched for this motif and could identify a possible AttR binding site 133 bases upstream of the fasA start codon (see Figure 14).

180 succ plant+succ gall+succ his+succ 160

140

120

100

80

60 GUS activity (u) activity GUS 40

20 0 D188 D188-5 D188-�att D188-�attR

Figure 13: Inducibility of translation of a fasA-uidA fusion on pIPRF1 in different genomic backgrounds grown under non-inducing (succ: succinate, plant+succ: plant extract/succinate) and inducing (gall+succ: gall extract/succinate, his+succ: histidine/succinate) conditions. Error bars represent SE (n=10).

GCGTTGCGCACAGGTGTCGAGGAACCATTCCTCCAGGGCTTCCGCGAACGCAGCATCGAC A L R T G V E E P F L Q G F R E R S I D mtr2

TACTTGGTTATCGCGTGCGAGCGTCTGTAGCGGTGGCCTGAGATATCCGAGTTATCCACG Y L V I A C E R L *

CGCCAGCGATCCTCGGCATCGGCGCGCCACTTCGTCCTTCTACAAACAATCGAAGGGATT

CCGGTATGGCCGGAACGGCTGATTTACCTTTAGAAATGCGACGCAACGGCCTGAACCCGA M A G T A D L P L E M R R N G L N P T fasA

Figure 14: DNA sequence and deduced amino acid sequence of the C-terminal part of mtr2 and the N-terminal part of fasA. The possible AttR binding site is indicated in bold, with the inverted repeats underlined and the typical T-N11-A bases marked in grey.

In D188-�att, where endogenous production of the autoregulatory compound and the feedback loop are absent, complementation with AttR restores inducibility of fasA to an intermediate level (see Figure 13), likely reflecting the activating capacity of the exogenously added inducer. To confirm this hypothesis, fasA expression levels were monitored in D188- �att[pIPRF1], incubated with different concentrations of gall extract. As shown in Figure 15A, fasA gene expression was activated at a certain threshold level of inducer present in the gall

Chapter 4 90

extract, but never reached the level of the wild type strain, due to the absence of a constant production of inducer molecules. The role of the feedback loop was demonstrated by doing a similar experiment with D188[pIPRF1]. At the lowest inducer concentrations, fasA expression increased gradually, but at a certain threshold level raised exponentially to reach almost immediately a maximum level. Possibly, this concentration of inducer reflects the minimal amount required for the activation of FasR and the subsequent signal transduction.

8 7 6 5 4 3 2 GUS activity activity (u) GUS 1 0 1 1,5 2 2,5 3,5 5,5 11 16 20,5 26 30 A gall extract (µL/mL

200 180 160 140 120 100 80 60 40 GUS activity activity (u) GUS 20 0 B 1 1,5 2 2,5 3,5 5,5 11 16 20,5 26 30 gall extract (µl/mL)

Figure 15: Translation of fasA in strains D188-�att[pIPRF1] (A) and D188[pIPRF1] (B) as a function of increasing amounts of gall extract.

The hyp locus of Rhodococcus fascians encodes a putative RNA helicase, but is not involved in fasA translation

The hyp locus was identified via insertional mutagenesis, yielding a hypervirulent strain D188-hyp7.1 (Crespi et al., 1992), and has been hypothesized to encode the translational repressor of fas expression (Temmerman, 2000). To evaluate this hypothesis, we first re- evaluated the sequence of the locus. Three ORFs were identified, the organisation of which was suggestive of an operon structure. The D188-hyp7.1 mutant carries an insertion in ORF1 (Temmerman, 2000), which likely results in a knockout of the complete operon (see Regulation and biochemistry of cytokinin biosynthesis in R. fascians 91

Figure 16). Whereas the G+C-content of ORF1 and 3 was in agreement with the average genomic G+C-content of R. fascians (61 to 68%), that of ORF2 was lower (see Figure 16). ORF1 and 3 were preceded by a putative ribosome-binding site. ORF1 possibly contains a transmembrane domain and ORF2 and ORF3 are predicted to be soluble.

1 2 3

ORF bp/aa Start Stop %GC Homology %id/%sim E- Reference value ORF1 660/220 ATG TGA 62.3 No significant homology (268) (927)

ORF2 270/89 ATG TGA 58.2 No significant homology (924) (1193)

ORF3 2558/852 ATG TGA 62.2 Putative helicase 65/76 0 Spatz et al., (Streptomyces 2002 (1200) (3757) violaceoruber)

Figure 16: Schematic representation of the hyp locus and sequence- and homology data of the genes. The insertion in D188-hyp7.1 is indicated by an arrow. Start: used start codon and position; Stop: used stop codon and position; bp/aa: nucleotide/amino acid length of the gene/gene product; %GC: GC content of the gene; %ID/%Sim: percentage identical/ similar amino acids on overall protein sequence; E-value: expectation value.

Similarity searches, done with the BLASTP program against the non-redundant protein database at NCBI, revealed no homologous sequences for ORF2 and ORF1, although the latter did contain an N-terminal Walker A motif involved in ATP/GTP binding. The ORF3 gene product was homologous to putative RNA helicases, which are RNA-dependent ATPases that catalyse ATP-driven conformational changes in RNA complexes and are involved in transcription, RNA decay and initiation of translation (Rocak and Linder, 2004; Rajkowitsch et al., 2007). The fas mRNA possibly folds into a secondary structure that captures the ribosome-binding site (Temmerman, 2000) and could be stabilised by the ORF3 protein thus preventing efficient translation. To evaluate the role of the hyp locus in fas gene expression, D188-�hyp, a mutant strain with a 6 kb deletion encompassing the entire hyp locus, was generated (Dreesden and Temmerman, unpublished results). Virulence assays on tobacco seedlings confirmed the hypervirulent phenotype of D188-hyp7.1 (Crespi et al., 1992) and revealed a similar

Chapter 4 92

phenotype for D188-�hyp: seedling growth was arrested at the cotyledonary stage and compared to infections with D188 more leafy galls developed (data not shown). We then introduced pIPRF1 into D188-hyp7.1 and D188-�hyp via electroporation and grew the recombinant strains under inducing and non-inducing conditions. Depending on the function of the hyp locus, three possible outcomes were predicted for these experiments. In case the locus encodes a repressor fasA expression in the mutants should be constitutive and high. In case it functions as an activator, no fasA expression should be measured. In case it is not involved at all, a wild type expression profile should be monitored. Figure 17 shows that fasA gene expression is induced in the hyp minus backgrounds to a wild type level, implying that the Hyp proteins are not involved in the regulation of fasA translation. Further experiments testing the inducibility of fasD, E and F in the hyp minus background are required however, to rule out that the putative hyp regulatory mechanism would be involved in regulating the translation of these genes.

250 succ succ+plant succ+gall succ+his

200

150

100

GUS activity (u) activity GUS 50

0 D188 D188-�hyp D188-hyp7.1

Figure 17: Inducibility of translation of a fasA-uidA fusion on pIPRF1 in different genomic backgrounds grown under non-inducing (succ: succinate, plant+succ: plant extract/succinate) and inducing (gall+succ: gall extract/succinate, his+succ: histidine/succinate) conditions. Error bars represent SD (n=2).

Discussion We have shown that R. fascians mainly produces iP, cZ and tZ and their 2MeS- derivatives (Chapter 3) most likely via the fas-encoded biosynthetic machinery. Based on the re-evaluation of the sequence of the fas locus, the biochemistry of FasD, FasE and FasF and the in vitro expression data, we propose the following biosynthetic pathway. The FasD gene encodes the essential Ipt, which can produce iP, but also tZ, when heterologously expressed in E. coli. Because it is currently unknown if the MEP-pathway, Regulation and biochemistry of cytokinin biosynthesis in R. fascians 93 required for the delivery of HMBDP as the donor for hydroxylated side chains (Krall et al., 2002; Rohmer, 2003), exists in R. fascians, we postulate that the main reaction product of FasD is iP. Expression of fasD is the highest amongst all genes of the locus, which supports its central position in the biosynthetic pathway delivering the precursor molecule for all other produced cytokinins. Subsequently, iP is hydroxylated to zeatin-type cytokinins by the P450- monooxygenase system (Takei et al., 2004), encoded by fasA, fasB and fasC, the only other functional module significantly expressed in vitro. It has been hypothesized that the energy required for the functioning of FasA would be generated by the reduction of pyruvate and the delivery of the electrons via ferredoxin. However, it cannot be ruled out that the C-terminal part of FasB together with FasC encode a true transketolase, which might contribute to the production of cytokinins by side chain delivery (Lange et al., 1998). Although the phosphoribohydrolase activity of FasF needs to be confirmed, this activity would potentially contribute to the level of produced cytokinins by releasing the free bases from their nucleotide precursors (Kurakawa et al., 2007). The low, in vitro expression of the gene possibly indicates that this function is only activated in planta. Similarly, the activities of mtr1, mtr2 and fasE seem to be triggered by another signal than that controlling fasA and fasD expression. Albeit that the enzymatic activities of Mtr1 and Mtr2 remain to be demonstrated, the 2- methylthiolation of tRNA-cytokinins by a SAM-dependent methyltransferase encoded by miaB (Pierrel et al., 2004), strongly suggests that mtr1 and mtr2 encode such activities. If so, these proteins would modify the zeatin-type cytokinins produced by the FasA-D enzymes to form 2MeScZ and 2MeStZ. The sequence of FasE suggested that the substrate specificity might be different from other CKX enzymes. FasE is indeed a functional cytokinin dehydrogenase with a somewhat unique action spectrum. iP and iPR are efficiently degraded by FasE and cZ, iP9N, 2MeScZ and 2MeStZ are not, which is comparable to the activities of apoplastic CKX enzymes of Arabidopsis (see Chapter 3). However, iPRMP and 2MeSiP are good substrates for FasE and tZ is not, which is very different for AtCKX (Galuszka et al., 2007). The presence of an active CKX in R. fascians is intriguing and seems contradictory to the essential role of cytokinins in the pathology. Possibly, FasE contributes to the balanced production of a specific subset of cytokinins in response to particular triggers that occur at defined stages of the interaction. The mechanisms regulating fas gene expression are very complex (Temmerman et al., 2000) and here we analysed the postulated role of AttR and the autoinduction of autoregulatory compound on fas gene translation (Maes et al., 2001; Cornelis et al., 2002). The fasA expression patterns obtained in D188-�att, D188-�att, and in these strains

Chapter 4 94

complemented with attR, demonstrated that fas gene induction is dependent both on a functional positive autoregulatory loop directing the production of high concentrations of the autoregulatory compound, and on AttR which was thus identified as the second transcriptional regulator controlling fas transcription. We also tested the hypothesized involvement of the hyp locus in fas translation (Crespi et al., 1992; Temmerman, 2000) by monitoring fasA expression in two hyp mutants, but our results showed that the Hyp proteins are not the translational regulators of fasA. Given the differential in vitro expression of the mtr and the fas genes, further experiments are required to rule out the possibility that the hyp- encoded regulatory machinery would be controlling translation of these genes. Assuming hyp is not involved in fas translation, we looked for other putative regulatory mechanisms. Virulence gene expression of Erwinia carotovora subsp. carotovora is very tightly controlled by environmental conditions, quorum sensing signals; an array of (post)transcriptional factors and the main regulator RsmA (Cui et al., 2008). RsmA is an RNA-binding protein that promotes RNA decay and suppresses exoenzyme production. Mutations in rsmA result in hypervirulence (Chatterjee et al., 1995). The similarities with players in fas gene regulation are striking (see Figure 18), nevertheless, no rsmA-like gene could be identified on pFiD188 (I. Francis, personal communication).

CarI Att/AttR

3-C6HL AC

Unknown R-protein FasR

RsmA unknown regulator

Exoenzymes Fas proteins

Figure 18: Similarities between virulence gene expression in E. carotovora (left) and R. fascians (right). Upon sensing the proper environmental conditions induction molecules are produced (3-C6HL or AC) which activate a transcription factor (unknown or FasR) which inhibits the expression of a translational repressor (RsmA or unknown) and allow expression of the virulence genes.

Another appealing regulatory mechanism is not mediated by regulating proteins, but acts via small molecules that bind to riboswitches, which are sensors for small ligands, such as amino acids and nucleobases. Upon ligand binding, these secondary structures in the mRNA undergo a conformational change resulting in accessibility of the ribosome bindings site (RBS) and translation initiation (Coppins et al., 2007). Putative secondary RNA structures encompassing the RBS are predicted upstream of fasA and fasD (Temmerman, 2000) and the autoregulatory compound (AC) produced by the Att proteins is a good potential candidate Regulation and biochemistry of cytokinin biosynthesis in R. fascians 95 as effector molecule. However, when the att and fas loci were introduced in D188-5, virulence was not restored (Crespi et al., 1994), indicating that other virulence factors are missing or that the fas genes were not correctly expressed. Although most riboswitches sense only one molecule, tandem riboswitch structures have been reported which sense two different molecules and only undergo rearrangement when both are present/absent (Coppins et al., 2007). Possibly, a second effector molecule besides (AC) could be necessary to induce a riboswitch conformational change allowing fas gene expression. A putative candidate is encoded by a locus located in unique region 1 of pFiD188, and situated between the att and hyp locus. This locus contains genes of which the encoded proteins have homology to biosynthetic enzymes involved in A-factor (2-isocapryloyl-3Rhydroxymethyl-�- butyrolactone) production, which are typical Gram-positive quorum sensing molecules that regulate expression of genes implicated in secondary metabolism and morphological differentiation (Nishida et al., 2007). Although it is currently unknown if this locus has a role in virulence, its expression is independent of AttR or the AC, but strictly dependent of FasR (our unpublished results), which would support the indirect role of FasR in fas translation (Temmerman et al., 2000). + + + AttR fas fasR att A-factor

+

+

+

Figure 19: New model for fas gene expression. Regulation of fas transcription is indicated in grey.

Assuming the A-factor locus has a role in fas gene expression, we postulate an adjusted working model (see Figure 19): as a result of autoinduction of att expression a sufficiently high concentration of AC is produced, which, together with the A-factor, is required to bind the riboswitches upstream of fasA and fasD, resulting in the release of the RBS and allowing translation. The biosynthesis of these molecules is controlled at the transcriptional level, respectively by AttR and FasR. Additional complexity in fas regulation arises since these two regulators also control fas gene transcription.

Chapter 4 96

Materials and methods

Bacterial strains, plasmids and growth conditions

The bacterial strains and plasmids used in this study are listed in tables 5 and 6. E. coli strains were grown at 37°C in Luria Broth (LB) medium (Sambrook et al., 1989), and R. fascians strains were grown at 28°C in Yeast Extract Broth (YEB) medium (Miller, 1972) for 2 days under gentle agitation until late exponential phase. When appropriate, media were supplemented with antibiotics: chloramphenicol (Cm) (25 µg/mL), phleomycin (Phleo) (1 µg/mL), carbenicillin (Cb) (200 µg/mL) or kanamycin (Km) (50 µg/mL). Plasmid isolation and DNA cloning were performed according to Sambrook et al. (1989) and R. fascians transformation was done as described before (Desomer et al., 1990).

Strains Description Reference

E. coli DH5� F-�80dlacZ�M15�(lacZ4A-argF)U169recA1endA1hsd Hanahan, 1983 R17(rK+mK+)supE44�- BL21 Star (DE3) F - ompT hsdS B (r B - m B - ) gal dcm rne131 Invitrogen

R. fascians D188 Wild type, virulent Desomer et al., 1988 D188-5 Plasmid-free, non pathogenic Desomer et al., 1988 D188-�att Deletion mutant of attRXABCDEFG, attenuated Maes et al., 2001 virulence, KmR D188-�attR Deletion mutant of attR, attenuated virulence, KmR Maes et al., 2001 R D188-hyp7.1 Insertion mutant in ORF1 of the hyp locus, Cm , Crespi et al., 1992 hypervirulent D188-�hyp Strain with a deletion of the 6a fragment of the linear Dreesden, plasmid. unpublished D188:mtr1 Integrated translational mtr1-uidA fusion, mtr1 insertion This work mutant, CmR D188:mtr2 Integrated translational mtr2-uidA fusion, mtr2 insertion This work mutant, CmR D188:fasA Integrated translational fasA-uidA fusion, CmR This work D188:fasD Integrated translational fasD-uidA fusion, CmR This work D188:fasE Integrated translational fasE-uidA fusion, fasE insertion This work mutant, CmR D188:fasF Integrated translational fasD-uidA fusion, CmR This work

Table 5: Bacterial strains

Regulation and biochemistry of cytokinin biosynthesis in R. fascians 97

Plasmid Marker Description Reference pDONR221 KmR Recombinatorial donor/master vector with Invitrogen M13 F and R primer sites pDEST17 CbR Recombinatorial cloning vector for N- Invitrogen terminal His-tagged protein expression under control of T7 promoter pDONR221fasD KmR fasD PCR fragment in pDONR221 This work pDEST17fasD CbR N-terminal fusion of fasD with a His-tag in This work the inducible expression vector pDEST17 pMalc2 CbR E. coli MBP overexpression vector New England Biolabs inducible by IPTG pMalc2fasE CbR Recombinant fasE-MBP overexpression Den Herder and construct Temmerman, 2002 pMalc2fasF CbR Recombinant fasF-MBP overexpression Den Herder and construct Temmerman, 2002 pRF37 Bifunctional cloning vector, capable of Desomer et al., 1990 replicating in E. coli as well in R. fascians pJDGV5 CmR CbR Replicating vector with a translational Temmerman et al., fasA-uidA fusion 2000 pIPRF1 PhleoR CbR Replicating vector with a translational This work fasA-uidA fusion and the attR gene with its own expression signals pSP72 CbR E. coli cloning vector Promega pGUS1 CbR Cloning vector for translational uidA Plant Genetic Systems fusions N.V. Belgium pIPmtr1 CmR CbR Cloning vector with a translational mtr1- This work uidA fusion pIPmtr2 CmR CbR Cloning vector with a translational mtr2- This work uidA fusion pIPfasA CmR CbR Cloning vector with a translational fasA- This work uidA fusion pIPfasD CmR CbR Cloning vector with a translational fasD - This work uidA fusion pIPfasE CmR CbR Cloning vector with a translational fasE - This work uidA fusion pIPfasF CmR CbR Cloning vector with a translational fasF - This work uidA fusion

Table 6: Plasmids

In silico analysis The program ORF Finder (NCBI) was used to detect the presence of open reading frames, which were translated to their corresponding amino acid sequences by using the Expasy translation tool. Homologous proteins were searched with the Basic Local Alignment Tool (BLAST) against the non-redundant protein database at NCBI. The program ClustalW was

Chapter 4 98

used to align sequences. The percentage GC was determined via the GC calculator program (http://www.genomicsplace.com/gc_calc.html). The program DAS (Dense Alignment Surface Plot) Transmembrane Prediction was used to detect transmembrane domains (Cserzo et al., 1997) and the Pfam database to detect motifs (Pfam 23.0 July 2008, 10340 families).

Generation of integrated translational fas-uidA in R. fascians.

Integrated translational fusions with the different fas genes were generated by single disruptive homologous recombination. The non-replicating plasmids pSPIPmtr1, pSPIPmtr2, pSPIPfasA, pSPIPfasD, pSPIPfasE and pSPIPfasF were introduced in D188 by electrotransformation, followed by selection on Cm. Appropriate insertion mutants were identified by Southern hybridisation with gene specific probes. Strains harbouring the linear plasmid with the desired single homologous recombination were used for expression analysis. The constructs for mtr1, mtr2 and fasE were made as such that their introduction into the target gene simultaneously generated a mutation. These mutants strains were assayed for their virulence in tobacco seedling assays (see Chapter 5).

Virulence tests

Nicotiana tabacum (L.) W38 seeds were sterilized by rinsing them for 2 min in 70% (v/v) EtOH, subsequently transferred to a 5% (w/v) NaOCl solution supplemented with 0.1% (v/v) Tween20, and washed with sterile water. The seeds were germinated and grown on full strength Murashige and Skoog medium in a growth chamber under a 16-h/8-h light/dark photoperiod at 24°C ± 2°C. After 5 days of germination, when the radicle emerged, a drop (± 20 µL) of a R. fascians culture, washed and concentrated in sterile water, was applied to the seedlings. Phenotypes (growth inhibition, thickening of the vascular tissue) were scored along the interaction.

Inductions and GUS assays

For the determination of fas gene expression, cultures were grown overnight under control and optimized conditions for virulence gene expression as previously described (Temmerman et al., 2000). A 2-days old R. fascians pre-culture was diluted 10 times in YEB medium. After overnight growth, the cells were washed and resuspended to OD600 2.0 in MinA medium (0.1%

NH4SO4, 0.025% MgSO4, 0.001% thiamine) adjusted to pH 5.0 in a citric acid/sodium citrate buffer system and supplemented with 20 mM succinate as carbon sources. Plant and gall extract were prepared by crushing plants or leafy galls, centrifuging and filtersterilising the Regulation and biochemistry of cytokinin biosynthesis in R. fascians 99 collected supernatant. As negative controls, nothing or 20 µL plant extract was supplemented, and for inducing conditions 5 mM histidine or 20 µL of gall extract was supplemented. After overnight incubation, the cells were collected by centrifugation, resuspended in 1 mL of MUG buffer (50 mM NaPO4, pH 7.0, 10 mM �-mercapto-ethanol, 10 mM EDTA, 0.1% SDS and

0.1% Triton X-100). Fifty µL was used to determine the OD600. The substrate 4- methylumbelliferyl-�-D-glucuronide (MUG) was added to a final concentration of 0.1 mM and incubated at 37°C for 1 hour, after which the reaction was stopped by adding a sample of 50

µL to 200 µL of 200 mM Na2CO3 in black nunc polysorb 96-well plates. GUS activity was determined by excitation at 365 nm and measurement at 460 nm (Fluostar optima reader) and is calculated as the measured emission/[time (min) * OD600].

Purification of recombinant FasD protein

A fasD expression vector was generated by using Gateway recombinant cloning technology. The coding region of fasD was amplified by PCR with gene specific primers flanked by attB recombination sites and cloned into the entry vector pDONR221 via att site recombination. The sequence of the gene was verified. Subsequently, it was cloned via att site recombination into the destination vector pDEST17 (Invitrogen) which expresses the N- terminal His-tagged recombinant protein and this construct was introduced into E. coli BL21 Star (DE3) (Invitrogen). Transformants were grown in LB medium, supplemented with

100 µg/mL carbenicillin and 1% (w/v) glucose at 37°C with shaking until OD600 = 0.6; after which FasD expression was induced by incubating with 0.4 mM IPTG at 20°C for 17 h. Cells were harvested by centrifugation (2000 x g, 30 min, 4°C) and pellets were resuspended in lysis/equilibration buffer (50 mM Tris/HCl pH 7.0; 300 mM NaCl, 10 mM imidazole, 10 mM �- mercapto-ethanol, 10 mM NaCl, 9% (w/v) glycerol). Lysozyme (0.75 mg/mL) and a protease inhibitor cocktail (Sigma-Aldrich) were added and the sample was incubated at 37°C for 15 minutes. All lysis and purification steps were performed at 4°C. The cell suspension was disrupted by a freeze-thaw procedure, after which the lysate was centrifuged for 20 min at 16100 x g and 4°C. The supernatant was applied on a Talon resin affinity column (Clontech Laboratories), which was previously equilibrated by lysis/equilibration buffer according to the Talon User Manual, and the loaded column was rotated in the fridge for 30 min. Subsequently, the column was washed twice with washing buffer (50 mM Tris, 300 mM NaCl, 10 mM imidazole, pH 7.0) and the sample was eluted with 3 mL elution buffer (250 mM imidazole, 300 mM NaCl, 50 mM Tris, 9% glycerol).

Chapter 4 100

All collected fractions (per mL) were E3 E2 E1 W2 W1 FT P M separated via SDS-PAGE (see Figure 20) showing that FasD (28 kDa) eluted in fractions E2 and E3, which were used for IPT activity measurement. The protein concentration was determined by using SDS-PAGE of the different the BCA Protein Assay (Bio-Rad) with Figure 20: purification steps for FasD-His (Silver bovine serum albumin as a standard. The staining). M: marker, P: pellet, FT: flow through, W1 and W2: wash fractions, average amount of eluted FasD protein E1,E2 and E3: elution fractions. was 0.2 mg/mL.

Isopentenyltransferase enzyme activity assay

All enzymatic assays were performed by adding 8 �g of purified protein to the reaction mixture (pH 8.0; 1 M betaine, 30 mM Tris, 50 mM KCl, 10 mM MgCl2, 5 mM �- mercaptoethanol). For the determination of the preferred side chain acceptor, the reaction mixture contained 25 �M DMAPP as donor and 0.001-3.2 �M AMP, ADP or ATP as substrate. To determine the preferred side chain donor, 2 �M AMP as acceptor and 0.001–1 �M DMAPP or HMBPP as donor were added to the reaction mixture. To determine pH optimum the enzymatic reaction was performed in citric acid-phosphate (pH from 3.5 to 7.5) or in Tris buffer (pH from 7.5 to 9) to which 50 mM KCl, 10 mM MgCl2, 5 mM mercaptoethanol, 0.08 �M DMAPP and 2 �M AMP were added. After incubating for 1 hour at 25°C, the reactions were stopped by heating the samples for 10 min at 96°C. Finally, samples were treated with calf intestine alcaline phosphatase (ClAP, Fermentas) for 30 min prior to estimation of the iPR concentration by ELISA. To estimate the FasD IPT activity, the iPR and ZR reaction products were quantified in all samples in flat bottom plates (MaxiSorp, Nunc Denmark) by using ELISA (modified from Strnad et al. 1990) with a specific anti-iPR and anti-ZR rabbit polyclonal antibody (Laboratory of Plant Growth Regulators, Olomouc). The wells of the 96-well plates were filled with 150 µL of 50 mM NaHCO3 (pH 9.6) containing 2 �L anti-IPR or 6 �L anti-ZR IgG respectively. The plates were incubated overnight at 4°C, and then washed twice with water to remove unbound antibody. In order to saturate the remaining protein binding sites, the wells were filled with 200 µL of 0.02% (w/v) BSA in Tris-buffered saline (TBS) buffer (50 mM Tris, 10 mM NaCl, 2 mM MgCl2, pH 7.5) and incubated for 60 min at 25°C. After being rinsed with water, the coated wells were filled with 100 �L of TBS, 50 �L of sample or control Ipt in TBS, and 50 �L of the cytokinin-alkaline phosphatase conjugate (0.1 g/mL) in 0.02% (w/v) BSA in Regulation and biochemistry of cytokinin biosynthesis in R. fascians 101

TBS (50 mM Tris, 10 mM NaCl, 2 mM MgCl2, pH 7.5) and the plates were incubated at 25°C for 60 min. Unbound conjugates were removed by rinsing the plates four times with TBS buffer. The activity of bound alkaline phosphatase was determined by incubating p- nitrophenylphosphate (150 �L, 1 mg/mL in 50 mM NaHCO3, pH 9.6) for 60 min at 25°C, after which the reaction was stopped by adding 25 �L of 0,5 M NaOH. The absorbance was measured at 405 nm and iPR and ZR concentrations were calculated based on a sigmoidal plot of their standard curves (see Figure 21).

0,45 0,7

0,40 0,6

0,35 0,5

0,30 0,4 0,25

A 405A 0,3 0,20 A405

0,2 0,15

0,10 0,1

0,05 0,1 1 10 0,1 1 10 iPR (pmol) ZR (pmol)

Figure 21: Standard curves for IPR or ZR quantification by ELISA.

Purification of recombinant FasE protein and cytokinin oxidase/dehydrogenase enzyme assay.

An E. coli pMalfasE preculture (Den Herder and Temmerman, +Xa –Xa M

2002) was diluted to OD600 0.1 and grown at 22°C until it reached OD600 0.5 and was induced overnight at 22°C with 1 mM IPTG. Subsequently, cells were collected and resuspended in 0.2 M Tris/HCL, pH 7.6, 5% glycerol, 200 mM NaCl, 1mM PMSF (60 mL) and disturbed on French press (20.000 psi). The lysate was centrifuged and the supernatant purified via amylose affinity chromatography. Fifteen mL of amylose resin (NEB) was equilibrated in 0.2 M Tris/HCl, pH 7.6, 5% glycerol, 200 mM NaCl, 1 mM PMSF. The sample was Figure 22: SDS-PAGE analysis of purified FasE- loaded onto the column and washed with 80 mL of MBP (-Xa) and treated with Factor Xa yielding FasE equilibration buffer and eluted with elution buffer (0.2 M (+Xa) (indicated by arrows). Tris/HCl, pH 7.6, 5% glycerol, 200 mM NaCl, 1 mM PMSF M: marker.

Chapter 4 102

supplemented with 20 mM maltose). The wash fraction was collected and applied again on the column, washed and eluted. The two elution samples were pooled and concentrated (MBP-RfCKX1 I) and purified once more as described above and concentrated.

Concentrated samples were supplemented with 10 mM CaCl2 and divided into 2 aliquots. One was treated with Factor Xa protease to yield free FasE (Sigma) (ratio 100 µg of protein 1 µg of protease) and incubated overnight at 22°C. Both samples (+Xa/-Xa) were checked via SDS-PAGE (see Figure 22) and tested for their CKX activity. The cytokinin dehydrogenase enzyme assay was performed as described before (Galuszka et al., 2007). The reaction mixture contained 100 mM McIlvaine buffer, pH 6.0, and 50 �M cytokinin substrate. The oxidase assay was performed as described by Frébort et al. (2002).

Purification of recombinant FasF protein and phosphoribohydrolase enzyme assay.

An E. coli pMalfasF preculture (Den Herder and NI I P S M E

Temmerman, 2002) was diluted to OD600 0.2 and grown at 37°C until it reached OD600 0.5 (NI) and was induced for 2h at 37°C by incubation with 0.3 mM IPTG (I) after which cells were collected via centrifugation. Cells were resuspended in column buffer (20 mM Tris-HCl pH 7.4, 200 mM NaCl, 1 mM dithiotreithol (DTT) and 1 mM EDTA) and disrupted via sonication (six 20 s pulses). The lysate was Figure 23: SDS-PAGE analysis of FasF-MBP overexpression and centrifuged and the supernatant (S) was purified via purification. NI: non-induced. I: induced, P: lysate amylose affinity chromatography. Amylose beads pellet, S: supernatant, M: marker, E: (New England Biolabs) were equilibrated in column eluted protein buffer. The sample was loaded onto the column, washed with column buffer and eluted with elution buffer (column buffer supplemented with 10 mM maltose). FasF-MBP (66kDa) purification was checked via SDS-PAGE (see Figure 23). The phosphoribohydrolase assay was performed as described by Kurakawa et al. (2007).

Author contributions.

IP gathered data for Figures 3-9, 11-17, 19, 23 and Table 1; and together with LS for Figures 10 and 20 and Table 2. LS gathered data for Figure 21. PG gathered data for Figure 22 and Tables 3 and 4. Chapter 5

�

A dedicated role for the Fas proteins in cytokinin production and virulence

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

“Biochemistry and biology of Rhodococcus� fascians cytokinin biosynthesis” (in preparation). � Pertry I, Václavíková K, Galuszka P, Spíchal L, Depuydt S, Temmerman W, Riefler M, Schmülling T, Strnad M, Holsters M, Tarkowski P, and Vereecke D.

A dedicated role for the Fas proteins in cytokinin production and virulence 105

Introduction

While de novo biosynthesis of iP and tZ has been well documented in bacteria and plants, biosynthesis of cZ and 2MeS-derivatives remains obscure (Sakakibara, 2006 and references therein). Despite the fact that in some plant tissues and species, such as chickpea seeds, rice and maize, cZ is the predominant zeatin-isomer (Emery et al., 1998; Murofushi et al., 1983; Veach et al., 2003), to date, no de novo cZ biosynthesis has been reported. These cytokinins are thought to be synthesised mainly as tRNA compounds, with their free base forms resulting from tRNA degradation, although another indirect biosynthesis pathway via the release from unidentified conjugates has been hypothesized (Miyawaki et al., 2006; Stirk et al., 2008). 2MeS-derivatives on the other hand, are thought to be solely synthesised as tRNA compounds (Prinsen et al., 1997; Anton et al., 2008). R. fascians tRNA contains iP, cZ, and 2MeScZ and was thought to be the primary source for the latter two, because the balances of the tRNA-bound and the free forms in the culture supernatant were relatively constant (Matsubara et al., 1968; Cherayil and Lipsett, 1977; Einset and Skoog, 1977; Murai et al., 1980). iP was hypothesised to be synthesised de novo, but no strong correlation could be made between the level of iP and the presence of fasD or the linear plasmid (Murai et al., 1980; Crespi et al., 1992). However, re-evaluation of the cytokinin profiles of D188 and D188-5, showed significant increases in the secreted levels of iP, cZ and 2MeScZ, which strongly suggest a linear plasmid-encoded de novo biosynthesis of these cytokinins (see Chapter 3). The only putative cytokinin biosynthesis genes on pFiD188 are encoded by the fas locus (I. Francis, personal communication) and the homologies of each fas gene are strongly suggestive for a specific enzymatic function in the cytokinin biosynthetic machinery. Moreover, the predicted enzymatic functions have been confirmed for FasD, FasE and FasF by in vitro biochemical analyses (see Chapter 4). Although FasD and FasA and/or FasB/C are essential for virulence, and FasE and/or FasF are required for full pathogenicity on older plants (Crespi et al., 1992; Crespi et al., 1994), the effect of these mutations on the cytokinin profile of R. fascians is currently unknown. Here, we first determined the spectrum of fas-dependent cytokinins using an activity- based approach on supernatants of fas mutant and overexpression strains. We then generated mutants in each predicted functional module of the biosynthetic machinery and thus, addressed the role of the different genes of the fas locus in virulence and in the produced cytokinin spectrum. To gain more insight in the mode of action of the cytokinin pool, we monitored the kinetics of in planta fas gene expression and of cytokinin perception and signal transduction during the interaction with Arabidopsis thaliana. Based on the data, we postulate a biosynthetic pathway and refine our model on the modus operandi of the bacterial cytokinins. Chapter 5 106

Results

The spectrum of cytokinins produced by R. fascians originates from the fas operon

To detect cytokinins or molecules with Figure 1: Schematic representation of the cytokinin activity present in fractionated R. complementation of the lethal sln1� phenotype by AHK4 and cytokinins (Inoue et fascians supernatant without relying on their al., 2001). structural features and because we know that Sln osmosensor AHK4 the R. fascians bio-active molecules are perceived by the cytokinin receptor AHK4 (see Chapter 3), we chose a cytokinin bio- assay described by Inoue et al. (2001) for the identification of fas-related molecules. This assay uses a Saccharomyces cerevisiae yeast mutant sln1� which is deficient in the sln1 gene, encoding an osmosensitive histidine kinase, and cannot grow in the absence of galactose because the response regulator, SSK1, is continuously de- phosphorylated, resulting in an overactivation of the downstream MAPK-pathway. Growth on galactose is possible through a galactose- inducible PTP2 phosphatase that blocks the MAPK-cascade. The Arabidopsis AHK4 cytokinin receptor gene, was introduced in this mutant, conferring cytokinin-dependent growth on glucose, because the cytokinins stimulate the histidine kinase activity of AHK4 which phosphorylates SSK1, blocking the overactivation of the MAPK-cascade and suppressing lethality of the sln1� mutation (see Figure 1). It was reported that iP, tZ, BAP and thidiazuron (TDZ) were effective in this assay (Inoue et al., 2001), but we decided to further explore the substrate specificity and sensitivity of this assay for cZ, DZ, 2MeSiP, 2MeStZ and 2MeScZ. Instead of performing the assay on solid medium (Inoue et al., 2001), the yeast was grown in liquid glucose medium to which cytokinins were added at different concentrations and after two days, the turbidity of the cultures was measured at OD600. As shown in Figure 2, all tested cytokinins were effective in rescuing the yeast although not to the same extent. cZ and DZ were only effective at very high concentrations, while iP, BAP, 2MeSiP, and 2MeStZ were still effective at 1 µM. tZ and 2MeScZ had an intermediate activity. We concluded that the broad substrate specificity of A dedicated role for the Fas proteins in cytokinin production and virulence 107

this assay was well suited to search for the cytokinins produced by the R. fascians fas- encoded machinery.

0,9 P 100 µM 10 µM 1 µM 0,8 0,7 0,6 0,5 0,4 OD600 0,3 0,2 0,1 0 Gal iP 2MeSiP cZ 2MeScZ tZ 2MeStZ BAP dZ

Figure 2: Growth of the yeast strain sln1� pCYC415AHK41 on galactose medium (Gal) as a positive control (P) or on glucose medium supplemented with different cytokinins at three concentrations.

To identify fas-dependent molecules, three strains were compared: the non-virulent fas1 mutant, which carries an insertion in the fasD gene on the linear plasmid (Crespi et al., 1992), the wild-type strain D188, and a fas overexpression strain D188-fasOE, which carries the fas operon under control of the stronger att promoter (Den Herder and Temmerman, 2002). Fifty mL cultures of these strains were grown under fas-inducing conditions (see Materials and Methods); their supernatant was collected, extracted and separated by HPLC under acidic conditions, as described by Temmerman (2000). Fractions were collected every minute and redissolved in 100 µL of sterile water or 80% MeOH and 10 µL was used for the yeast bioassay. With this activity-based approach two fas-dependent cytokinin-active fractions, 17 and 25, were identified (see Figure 3C, D and E). Fraction 17 co-elutes with iP and BAP (around 16,5 and 17 minutes), and fraction 25 co-elutes with 2MeSiP (around 24.5 minutes), both were bioactive in all three strains although the level of yeast growth is clearly dependent on the presence and expression level of the fasD/fas operon genes. However, no distinctive peaks in the chromatograms seemed to correspond to the bioactive fractions, indicating that the concentrations are very low (see Figure 3). After these analytical experiments, a large-scale isolation procedure was initiated for 35 L of cultures and the obtained active fractions of the acidic HPLC were pooled and subjected to a second HPLC fractionation under alkaline conditions, as described by Temmerman (2000), to obtain purer molecules for structure determination. This second purification of bioactive fraction 17, revealed a fas-upregulated peak, co-eluting with iP around 18 minutes (see figure 4A and B). Fractions 17.18 and 17.19, corresponding to this peak, consistently showed cytokinin activity for all preparations and for the 3 strains (data not shown). Chapter 5 108

Although less pronounced, the second HPLC purification of fraction 25 also revealed a fas-upregulated peak, co-eluting with 2MeSiP around 24.50 minutes (see figure 4A and C). Two bioactive fractions were correlated with this peak: fraction 25.24, which was fas- dependent, and fraction 25.25, which was present in all 3 strains and was fas-upregulated (data not shown).

2MeSZR Z

BAP 2MeSiP

iP TDZ

A 17 25 D188-fas1

D188

D188-fasOE

B 1 1 1 9 9 9

C D E

Figure 3: (A) Chromatogram showing the fractionation of standard cytokinins detected at 234 nm. The elution time in the X-axis is indicated in minutes. (B) Chromatogram comparing the fractionation of supernatant extract of D188-fas1, D188 and D188-fasOE visualised at 234 nm. The elution time is indicated in the X-axis in minutes and the active fractions are indicated. Yeast bioassay with D188- fas1 (C), D188 (D) and D188-fasOE fractions (E). A dedicated role for the Fas proteins in cytokinin production and virulence 109

A D188-fas1

D188

D188-fasOE

B

D188-fas1

D188

D188-fasOE

C

Figure 4: (A) Chromatogram showing the fractionation of standard cytokinins visualised at 234 nm. Chromatogram of the fractionated bioactive fractions 17 (B) and 25 (C) of D188-fas1, D188 and D188- fasOE visualised at 268 nm. An arrow indicates the peaks correlated with bioactive fractions. The elution time in the X-axis is indicated in minutes. Chapter 5 110

The bioactive fractions isolated with this procedure, were further purified by immuno- affinity chromatography allowing separation of isoprenoid, aromatic, and highly substituted cytokinins, their free bases as well as their ribosides and glucosides. Since it was unknown whether the fas-dependent cytokinins had classical cytokinin structures, the eluate as well as the flow-through were recovered and tested in the yeast bioassay. The flow throughs did not exhibit cytokinin activity (data not shown), indicating that the active compounds in the fractions do have classical cytokinin structures. Based on the activity in the eluates (see Figure 5), it seems that fraction 25.24 was fas-dependent, fractions 25.25 and 17.19 were fas-enhanced, and fraction 17.18 was similar for the three bacterial strains. Subsequent identification of the cytokinins in these fractions by Q-TOF micro-mass spectrometer analysis, showed that fractions 17.18 and 17.19 contained iP, whereas fractions 25.24 and 25.25 contained 2MeSiP and 2MeScZ. These results confirm that R. fascians has (a) chromosomal pathway(s) that produce(s) iP, 2MeSiP and 2MeScZ, and indicate that the fas locus directs the increased biosynthesis of the same molecules. It should be noted however, that the followed procedure only allows the identification of the most abundant and/or easily ionized derivatives.

1.000 0.800

0.800 0.600 0.600 600

600 0.400

OD 0.400 OD

0.200 0.200

0.000 pos.control D188-fas117.18 D18817.18 D188-fasOE17.18 D188-fas117.19 D18817.19 D188-fasOE17.19 0.000 pos.control D188-fas125.24 D18825.24 D188-fasOE25.24 D188-fas125.25 D18825.25 D188-fasOE25.25

A B

Figure 5: Yeast growth obtained with immuno-affinity eluates of bioactive HPLC fractions 17 (A) and 25 (B) from induced supernatant of R. fascians strains D188-fas1, D188 and D188-fasOE. The positive (pos.) control was grown on galactose medium.

We most likely did not detect cZ as a fas-dependent cytokinin in this analysis because the yeast bioassay is not very sensitive for this molecule; tZ and 2MeStZ are produced in too low quantities to be detected by this bioassay.

A dedicated role for the Fas proteins in cytokinin production and virulence 111

The fas mutants have different cytokinin profiles

To address the role of each of the proteins encoded by the fas locus in bacterial cytokinin production, different fas insertion mutants were generated via disruptive single homologous recombination as described in Chapter 4 (see Materials and methods for more details). These recombinant strains, D188-fas1 and D188-fas6 (see Figure 6) were grown for 17 hours under inducing conditions with succinate as a carbon source, and the cytokinin levels in their supernatant were profiled by LC-MS.

1 2 3 4 5 6

Figure 6: Schematic representation of the fas locus and the different fas insertion mutants used for the cytokinin profiling and virulence assays. 1: D188-mtr1; 2: D188-mtr2; 3: D188-fas6; 4: D188-fas1; 5: D188-fasE and 6: D188-�fasF.

Interestingly, all mutations in the fas locus resulted in higher levels of 2MeSiP compared to those in D188 and D188-5 (see Figure 7), confirming that 2MeSiP biosynthesis is not fas but chromosome-dependent (see Chapter 3). This pattern may indicate that the chromosomal and the Fas-pathway compete with each other for precursors. Because D188- 5 (fas- pFiD188-) does not have higher 2MeSiP levels, these precursors would have to be delivered by linear-plasmid encoded proteins or the linear plasmid would stimulate the precursor production by the chromosome. The latter seems more likely, since analysis of the coding potential of pFiD188 did not reveal proteins potentially involved in the synthesis of DMAPP or nucleotides (I. Francis, personal communication). The higher levels in all fas insertion mutants (fas- pFiD188+) could also imply that chromosomally produced 2MeSiP is used as a precursor by the Fas-machinery to make other cytokinins. The strong increase in 2MeSiP in the fasA mutant could indeed indicate that chromosomal 2MeSiP is hydroxylated by FasA to form 2MeSZ. Whereas 2MeStZ had been detected in supernatant of D188 only (see Chapter 3), its 2MeStZ level dropped below detection limits in all mutants tested, demonstrating that the entire fas locus is involved in 2MeStZ production (data not shown). The levels of the other cytokinins (iP, cZ, 2MeScZ and tZ) were significantly altered in the different mutants, albeit to a different extent (see Figure 7), indicating that each Fas protein is involved in a different step of the biosynthetic pathway. Chapter 5 112

The central position of FasD in the biosynthetic machinery was underlined by the reduction of cZ, 2MeScZ and tZ to D188-5 levels, and proves that the Ipt directs the first step for the production of a cytokinin spectrum by delivering iP, as a precursor for the other Fas proteins. In D188-fas1 iP levels were also significantly lower than in D188, nevertheless, they were 3-fold higher than in D188-5, again suggesting a possible competition between Fas and the chromosome and a stimulatory role of pFiD188 on chromosomal iP production (see Figure 7).

11 1,4 10 1,2 9 8 1 7 6 0,8 nM 5 nM 0,6 4 3 0,4 2 0,2 1 0 0 iP 2MeSiP

3 5,5 5 2,5 4,5 2 4 3,5 1,5 3 nM nM 2,5 1 2 1,5 0,5 1 0,5 0 0 cZ 2MeScZ 0,045 0,04 0,035 0,03

nM 0,025 0,02 0,015 0,01 0,005 0 tZ

Figure 7: Cytokinin production by different R. fascians strains grown under inducing conditions with succinate as carbon source. Error bars represent SDs (n=3). D188: wild type; D188-5 linear plasmid-free; D188-mtr1; D188-mtr2; P450: D188-fas6; IPT: D188-fas1; CKX: D188-fasE and LDC: D188-�fasF. A dedicated role for the Fas proteins in cytokinin production and virulence 113

The cytokinin profiles identified the putative FasA-C hydroxylating system as equally important, because also in the D188-fas6 mutant, tZ, cZ and 2MeScZ levels were brought back to those measured in D188-5. The level of iP increased with over 50%, confirming that the fasA mutation has no polar effect on fasD (Crespi et al., 1994) and indicating that iP is a precursor for FasA-mediated hydroxylation to zeatin-type cytokinins (see Figure 7). In Chapter 4, we demonstrated that FasE efficiently degraded iP and 2MeSiP. These data are supported by the cytokinin profile of D188-fasE: iP levels were doubled, while 2MeScZ levels were unaltered. Surprisingly, this mutation also resulted in a decreased level of cZ and tZ, similar to that in D188-5. However, the cZ, 2MeScZ and tZ levels are comparable for D188-fasE and D188-�fasF, suggesting that the mutation in fasE has a polar effect on fasF. The strongly reduced levels of zeatin-type cytokinins in D188-�fasF further imply that the putative phosphoribohydrolase activity of FasF (see Chapter 4) is involved in the release of cZ and tZ, and to a lesser extent of 2MeScZ from their respective nucleotides. Unexpectedly, Mtr1 and Mtr2 do not appear to be involved in the methylthiolation of iP and cZ. A possible role in the modification of tZ could not be evaluated because the levels of 2MeStZ were below detection limits. Nevertheless, the different cytokinin profiles of the two mtr mutants show that the mutation in mtr1 has no polar effect on mtr2, that Mtr1 modifies iP, and that both proteins are involved in tZ and cZ production (see Figure 7). The biochemical activity of Mtr1 and Mtr2 currently remains elusive. Most importantly, all fas genes strongly affect the production of tZ and cZ, demonstrating that these are the main fas cytokinins. Moreover, the absolute dependence of cZ levels on the fas locus, and of 2MeScZ levels on FasA and FasD are an important indication that these cytokinins are synthesised de novo.

The Fas proteins are differentially involved in virulence

We then analysed the correlation of the different cytokinin profiles of the six insertion mutants on their capacity to inhibit tobacco seedling growth. Typically, symptoms induced in this assay can be divided into three categories: no growth inhibition and no development of malformations (NV, non-virulent), intermediate growth inhibition with malformations and development of few leaves (I, intermediate), and complete growth inhibition and arrest at the cotyledonary stage (V, virulent). Infection with D188 resulted in almost a complete growth inhibition, although with some escapes (see Figures 8A, 9 and 10), while infection with D188-5 did not affect seedling growth (see Figures 8B and 9). Strains D188-fas1 and D188-fas6 were non-virulent (see Figures 8C and D), confirming the absolute requirement of FasA and FasD for virulence Chapter 5 114

(Crespi et al., 1992; 1994). Similarly, both mtr mutants were not capable of inhibiting seedling growth, demonstrating that they are also essential for virulence (see Figures 8E and F and Figure 9). Because the cytokinin profiling indicated that the mutation in mtr1 had no polar effect on mtr2, despite their strong similarity (see Chapter 4), their non-pathogenic phenotype implies that they are apparently not functionally redundant. The mutation of fasE generated an attenuated phenotype (see Figures 8G and 9). The majority of the seedlings developed normally, some seedlings were completely arrested and did not resume growth, but seedlings with an intermediate growth inhibition did grow out. These results suggest that bacterial CKX activity is required both for establishing and maintaining symptoms. The D188-�fasF mutant had a very interesting phenotype. At first, this strain appeared to be almost fully virulent (see Figures 8H and 10A). However, as the interaction proceeded, symptoms were not maintained: growth inhibition was lost, and the seedlings resumed growth (see Figures 9 and 10A-E), indicating that FasF is not required for symptom initiation, but is essential for symptom maintenance. Infection of Arabidopsis with the different fas mutants yielded comparable phenotypical profiles (data not shown).

A B C D

E F G H

Figure 8: Phenotypes of tobacco seedlings inoculated with different R. fascians strains at 17 dpi. A: D188, B: D188-5, C: D188-fas1 (fasD), D: D188-fas6 (fasA), E: D188:mtr1; F: D188:mtr2, G: D188:fasE ; H: D188-�fasF.

A dedicated role for the Fas proteins in cytokinin production and virulence 115

V I NV

D188 D188-5 mtr1 mtr2 CKX LDC 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 24 45 24 45 24 45 24 45 24 45 24 45 days post infection

Figure 9: Phenotype of tobacco seedlings inoculated with different R. fascians strains at 24 and 45 days post infection. At least 50 plants were scored. V: virulent; I: intermediate growth inhibition, NV: non-virulent. D188: wild type; D188-5: linear plasmid- free derivative; mtr1: D188:mtr1; mtr2: D188:mtr2; CKX: D188:fasE ; LDC: D188-�fasF.

A B C

D E

Figure 10: Phenotypes of tobacco seedlings inoculated with R. fascians D188 (right part of each plate) and D188-�fasF (left part of each plate). A: 14 dpi; B: 28 dpi; D: 28 dpi with large plants removed; C: 39 dpi; E: 39 dpi with large plants removed.

Chapter 5 116

The virulence assays demonstrate that the Fas proteins have a differential role in symptom development, likely by producing different types or concentrations of cytokinins as suggested by their cytokinin profiles. A possible mechanism to accomplish differential cytokinin production could be the establishment of different expression kinetics. Our data in Chapter 4 already indicated that different regulatory mechanisms control the expression of the fas genes in vitro, and so we infected tobacco seedlings with strains carrying different translational fas-uidA fusions described in Chapter 4, and monitored the in planta expression of mtr1, mtr2, fasA, fasD, fasE and fasF 1, 3, 13, 21, and 28 dpi. The experiments were repeated 3 times and although expression levels were highly variable, some patterns could be discerned: fasF expression was low throughout the interaction, mtr1 and mtr2 expression peaked around 13 days and expression of fasA, D and and E peaked between 13 and 21 days depending of the experiment (data not shown).

Perception of R. fascians cytokinins by and signal transduction through AHK4 is not maintained throughout the interaction

In Chapter 3, we postulated that ectopic AHK4 expression increases sensitivity of the plant towards R. fascians signals. To evaluate the kinetics of this process, we followed ARR5:GUS expression as a measure of the cytokinin response in Arabidopsis plants in which AHK4 is the only active cytokinin receptor (ahk2 ahk3 double mutant). As a control, we monitored ARR5:GUS expression in wild type plants and in Arabidopsis plants in which AHK3 is the only active cytokinin receptor (ahk4 ahk2 double mutant) (see Figure 11). In wild type plants, ARR5 expression was low and restricted to the vasculature of the shoot apical meristem (SAM). Upon R. fascians infection, a transient expansion of the expression throughout the SAM, including leaf primordia was observed at 4dpi (see figure 11) (Depuydt, 2008). Uninfected cytokinin receptor mutants exhibited an expanded ARR5 expression (see Figure 11), which is likely correlated with their higher cytokinin levels resulting from a malfunctioning feedback system. In wild type plants, upon cytokinin perception, a feedback mechanism is activated to maintain an optimal cytokinin ratio. When this feedback is disturbed, homeostasis can be reduced, but mainly cytokinin biosynthesis is increased resulting in a stronger signal transduction (Riefler et al., 2006). A dedicated role for the Fas proteins in cytokinin production and virulence 117

D188. D188.

R. fascians R. wild type ColO and double cytokinin receptor mutant lines mock-inoculated with water (control) or (control) water with mock-inoculated lines mutant receptor cytokinin double and ColO type wild of ARR5:GUS analysis Histochemical

5:GUS 5:GUS 5:GUS 5:GUS 5:GUS 5:GUS +D188 +D188 +D188 +D188 +D188 +D188 control control control control ahk2 ahk3 ahk3 ahk2 ahk2 ahk3 ahk3 ahk2 ahk2 ahk4 ahk2 ahk4 ARR ARR ARR ARR ARR ARR infected with infected Figure 11: Figure

Chapter 5 118

Upon infection, we observed a very strong increase of ARR5 expression for both receptor mutant lines from 4dpi onwards (see Figure 11). However, whereas expression remained high throughout the experiment when AHK3 was the only functional cytokinin receptor, expression was transient when only AHK4 was functional and started to decrease from 17 dpi on. These results indicate that AHK4-mediated signal perception and transduction largely coincides with strong fas gene expression, implying that it is involved in intensifying sensitivity at the initiation phase of symptom development. AHK3, on the other hand, is involved throughout the interaction

Discussion

Cytokinin profiling of the supernatant of R. fascians cultures identified 6 types of cytokinins - iP, cZ and tZ and their 2MeS-derivatives - and except for 2MeSiP, the linear plasmid contributed to their production (see Chapter 3). The identification of bioactive fractions derived from D188, the ipt mutant D188-fas1, and a fas overexpression strain D188-fasOE demonstrated that the fas operon indeed contributed to iP and 2MeScZ production. Its role in tZ, 2MeStZ, and cZ could not be confirmed with this approach because the produced levels were too low and/or the yeast bioassay was not sensitive enough. Analyses of the cytokinin profiles and the virulence phenotypes of the different fas mutants, revealed that each Fas protein had a dedicated role in cytokinin production and symptom development. Based on the results we postulate a model for the fas-encoded cytokinin biosynthetic machinery (see Figure 12). FasD is the key enzyme for virulence- associated cytokinin production by synthesizing iP, which serves as a precursor for the other cytokinins and thus is essential for virulence. cZ, tZ and 2MeStZ production requires the entire fas locus, and these cytokinins are therefore unambiguously identified as the main fas cytokinins. FasA, as a putative P450 mono-oxygenase hydroxylates iP to form zeatin-type cytokinins. In addition, FasA also hydroxylates chromosomally produced 2MeSiP, producing 2MeScZ. The central position of FasA in the formation of all zeatin-type cytokinins is in agreement with its requirement for virulence. FasF has phosphoribohydrolase activity, which releases tZ, 2MeStZ, cZ, and to a lesser extent 2MeScZ from their nucleotide precursors. This alternative production of these cytokinins is not essential for symptom induction, but is absolutely required for symptom maintenance. The role of fasE in the determination of the cytokinin spectrum and virulence is more difficult to interprete: although the D188:fasE mutant seems to produce more cytokinins than D188-�fasF, it is less virulent. Similarly, the mtr mutants are less affected in the production of the six cytokinins than D188- �fasF, yet A dedicated role for the Fas proteins in cytokinin production and virulence 119

they are not virulent. With the collection of fas mutants generated during this study, we could initiate a broader metabolic profiling, which might clarify these points.

ADP (/AMP/ATP) + DMAPP Chromosome FasD FasE adenine + 3-methyl-2-butenal iPRMP 2MeSiP iP FasA FasA-C FasF FasA + FasF + Mtr1 & 2 2MeScZ

tZ, 2MeStZ and cZ

Figure 12: Model on Fas-mediated cytokinin biosynthesis in R. fascians D188.

When in planta expression of the fas genes was evaluated, the results were highly variable. Nevertheless, generally, expression of the fas genes was highest in the initial phase of the interaction. When symptoms were fully established, their expression decreased to a basal level. These kinetics imply that symptom initiation requires a high level of fas- cytokinins, whereas for symptom maintenance a much lower level is sufficient. Unfortunately, because of the variable results, no correlation could be made between the timing of expression onset and virulence phenotypes. Nevertheless, AHK4-mediated induced ARR5 expression upon D188 infection did exhibit similar kinetics as in planta fas gene expression, suggesting that AHK4 perception of the fas cytokinins is only relevant for the establishment of symptom development and not during the symptom maintenance. In agreement with such a function, transient upregulation of AHK4 is coinciding with the timing of shoot commitment during regeneration from root explants (Che et al., 2002). It is tempting to assume that the cytokinins produced via the FasF phosphoribohydrolase, which are supposed to be involved in symptom maintenance, are not well perceived by AHK4. Indeed, AHK4 is thought to be highly specific and mainly recognizes iP and tZ, and the ARR5 expression in the AHK4 background reflects a shift in cytokinins produced by the bacteria. AHK3 is expressed throughout the interaction and recognizes a broader cytokinin spectrum, including ribosides, ribotides, cZ and DZ (Spichal et al., 2004; Yamada et al., 2001; Romanov et al., 2006) constitutively. In conclusion, we have shown that the fas locus is indeed responsible for the pFiD188- dependent enhancement of the cytokinin spectrum. Moreover, the different fas genes seem to Chapter 5 120

have specific roles in the biosynthetic pathway. The virulence phenotypes of the fas mutants suggest that the different cytokinins in the spectrum or their relative concentrations differentially mediate symptom initiation and maintenance. The occurrence of such cytokinin waves is supported by the transient activation of AHK4-mediated cytokinin signal transduction.

Materials and Methods

Bacterial strains, growth conditions and plasmids. The bacterial strains and plasmids used in this study are listed in tables 1 and 2. E. coli strains were grown at 37°C in Luria Broth (LB) medium (Sambrook et al., 1989), and R. fascians strains were grown at 28°C in Yeast Extract Broth (YEB) medium (Miller, 1972) for 2 days under gentle agitation until late exponential phase. When appropriate, media were supplemented with antibiotics: chloramphenicol (Cm) (25 µg/mL), phleomycin (Phleo) (1 µg/mL), carbenicillin (Cb) (200 µg/mL) or kanamycin (Km) (50 µg/mL). Plasmid isolation and DNA cloning were performed according to Sambrook et al. (1989) and R. fascians transformation was done as described before (Desomer et al., 1990).

Strains Description Reference

E. coli DH5� F-�80dlacZ�M15�(lacZ4A-argF)U169recA1endA1hsd Hanahan, 1983 R17(rK+mK+)supE44�-

R. fascians D188 Wild type, virulent Desomer et al., 1988 D188-5 Plasmid-free, non pathogenic Desomer et al., 1988 D188-fas1 Insertion mutant in fasD, avirulent, CmR Crespi et al., 1992 D188-fas6 Insertion mutant in fasA, avirulent, CmR Crespi et al., 1994 R D188-fas5 Insertion mutant in fasE, attenuated virulence, Cm Crespi et al., 1994

D188-fasOE fas5 with the fas overexpression construct pRF37(attP::fasA- This work F) D188:mtr1 Integrated translational mtr1-uidA fusion, mtr1 insertion This work mutant, CmR D188:mtr2 Integrated translational mtr2-uidA fusion, mtr2 insertion This work mutant, CmR D188:fasA Integrated translational fasA-uidA fusion, CmR This work D188:fasD Integrated translational fasD-uidA fusion, CmR This work D188:fasE Integrated translational fasE-uidA fusion, fasE insertion This work mutant, CmR D188:fasF Integrated translational fasD-uidA fusion, CmR This work D188-�fasF fasF insertion mutant, KmR This work

Table 1: Bacterial strains A dedicated role for the Fas proteins in cytokinin production and virulence 121

Plasmid Marker Description Reference pCRBlunt KmR E. coli cloning vector for cloning of Invitrogen blunt-end PCR products pCRBlunt-�fasF KmR pCRBlunt with an fasF internal PCR This work fragment R pRF37(attP::fasA-F) Phleo , Replicating vector with fasA-F under Den Herder and CbR control of the att promoter Temmerman, 2002 pSP72 CbR E. coli cloning vector Promega pGUS1 CbR Cloning vector for translational uidA Plant Genetic Systems fusions N.V. Belgium pIPmtr1 CmR CbR Translational mtr1-uidA fusion This work pIPmtr2 CmR CbR Translational mtr2-uidA fusion This work pIPfasA CmR CbR Translational fasA-uidA fusion This work pIPfasD CmR CbR Translational fasD -uidA fusion This work pIPfasE CmR CbR Translational fasE -uidA fusion This work pIPfasF CmR CbR Translational fasF -uidA fusion This work

Table 2: Plasmids

Inductions of R. fascians cultures.

To induce virulence gene expression, R. fascians cultures were grown as previously described (Temmerman et al., 2000). A 2-days old R. fascians pre-culture was diluted 10 times in YEB medium. After overnight growth, the cells were washed and resuspended to OD600 2.0 in MinA medium (0.1% NH4SO4, 0.025% MgSO4, 0.001% thiamine) adjusted to pH 5.0 by using citric acid and sodium citrate as a buffer system, supplemented with 20 mM succinate as a carbon source and 5 mM histidine.

High Pressure Liquid Chromatography (HPLC) fractionation and immuno-purification.

For HPLC fractionation 500 mL cell-free cultures were acidified using trifluoroacetic acid (TFA) to a final concentration of 0,1% and extracted by passing them over C18 cartridges (Sep-Pak, Waters), washed with 0,01% TFA and eluted with 80% methanol. The eluates were dried in vacuo and resuspended in 500 µL 0,1%TFA; 2% acetonitrile and filtered by passing them over Millex®-HV filter units (0.45 µm). Samples were subsequently analysed using an HPLC system (Waters 625 LC System) connected to a photodiode array detector (Waters™ 996). The samples were fractionated using a reverse-phase column (4 x 250 mm, Vydac, 218 TP, C18, 300 Å, with guard column, Alltech) with water (A) and acetonitrile (B) as mobile phases. A linear gradient was set from 2% B to 50% B in 30 minutes and 50% to 100% B in 5 minutes, followed by a regeneration of 15 minutes in 2% B. The flow rate was 1mL/min and eluate was Chapter 5 122

collected every minute. The collected fractions were dried in vacuo and tested for cytokinin activity in the yeast bioassay described below. Subsequently, the fractions exhibiting cytokinin activity were pooled and submitted to a second separation using the same HPLC system and reverse-phase column (4 x 250 mm, Vydac, 218 TP, C18, 300 Å, with guard column, Alltech) with 10 mM ammonium acetate (A) and 50/50 (v/v) 10 mM ammonium acetate/acetonitrile (B). The gradient was set from 10% B to 90% B in 20 minutes, followed by a regeneration of 10 minutes in 10% B. Again the flow rate was 1mL/min and eluate was collected every minute. The collected fractions were dried in vacuo and tested for cytokinin activity in the yeast bioassay described below. The samples were further purified via immuno-purification according to Novak et al. (2003) (Laboratory of Plant growth regulators, Palacky University, Olomouc, Czech Republic).

Yeast bioassay.

The yeast strain S. cerevisiae sln1� carrying the plasmid p415Cyc1CRE1 (Inoue et al., 2001; obtained from the lab of Kakimoto) was grown at 28°C on Yeast nitrogen base (YNB, 6,7g/L) medium supplemented with 2% galactose or and 0,65 g DO supplement -Ura/-His/-Leu and agar, to which 8mL 100mM His was added after autoclaving. Fifty µL of a water suspension containing S. cerevisiae sln1� [p415Cyc1CRE1] was diluted in 20 mL YNB medium supplemented with 2% glucose or and 0,65 g DO supplement -Ura/- His/-Leu, to which 8mL 100mM His was added after autoclaving and 190 µL added to a 48- well plate. To test the fractions for their cytokinin activity, they were resuspended in 100 µL of water and 10 µL was added to each well. As a negative control water was added and as a positive control 10 µM tZ. The plates were incubated at 28°C and growth was scored after two days.

Identification of cytokinins in bio-active fractions.

A hybrid Q-Tof micro mass spectrometer (Waters MS Technologies, Manchester, UK) was used for the high-resolution identification the cytokinins in the bio-active fractions. Aliquots of 5 �l were taken for analysis. The measurement was performed in connection with HPLC analysis on a CapLC® separation module (Waters, Milford, MA, USA) using a reversed- phase column Symmetry C18 (300 µm x 150 mm, 5 �m; Waters). Following the injection, analytes were eluted with a 60-min binary linear gradient (0-1 min, 30% B; 1-20 min, 50% B; 20-36 min, 100% B; 36-45 min, 100% B; 45-60 min, 30% B; flow-rate of 7 µl/min; column temperature of 30°C) of 15mM HCOONH4 pH4 (A) and methanol (B). Electrospray ionization in the positive ion mode was performed using the following parameters: source A dedicated role for the Fas proteins in cytokinin production and virulence 123

block/desolvation temperature, 90°C/200°C; capillary/cone voltage, 2,500/28 V; and spray/cone gas flow (N2), 50/200 l/h. In the full-scan mode, data were acquired in the mass range m/z 50 to 1000, with a pusher cycle time of 33 ms, a scan time of 2.0 s, and collision energy of 4 V. For the exact mass determination experiments, a lock spray was used for external calibration with a mixture of 0.1 M NaOH/10% formic acid (v/v) and acetonitrile (1:1:8 by volume) as a reference. Accurate masses were calculated and used for the determination of the elementary composition of the analytes with a fidelity of 5 ppm.

Insertion mutagenesis of R. fascians.

Each fas insertion mutant and/or uidA strain was isolated via single disruptive homologous recombination. Insertion mutants in the different fas genes of the linear plasmid were obtained by electrotransformation of D188 with the non-replicating plasmids pSPIPmtr1, pSPIPmtr2, pSPIPfasA, pSPIPfasD, pSPIPfasE, pSPIPfasF and pCR�fasF followed by selection on the appropriate antibiotic. Appropriate insertion mutants were identified by Southern hybridisation with gene specific probes. Strains harbouring the linear plasmid with the desired single homologous recombination were assayed for expression and virulence, and their cytokinin levels were profiled. The constructs for mtr1, mtr2 and fasE were made as such that their introduction into the target gene simultaneously generated a mutation and a translational uidA fusion.

Bacterial cytokinin measurements.

For measurement of classic cytokinin levels, cytokinins were isolated, purified and analyzed according to Novák et al. (2003) with some modifications. 2MeS-cytokinins were quantified by LC-MS in multiple reaction-monitoring mode. For cytokinin profiling of bacterial supernatants, 300 mL cultures were grown for 17 hours under control and optimized conditions for virulence gene expression as described above. Subsequently, 100 mL cell-free cultures were extracted by passing them over C18 cartridges (Accubond, Agilent Technologies), which were washed with water and eluted with C18 elution buffer (80% methanol, 2% acetic acid) and used for two technical replicates; these analyses were independently repeated 3 times. The occurrence of 25 cytokinin metabolites was evaluated, including isoprenoid and aromatic bases, their ribosides, and O- and N-glucosides.

Plant material and growth conditions.

A. thaliana and Nicotiana tabacum (L.) W38 seeds were sterilized by rinsing them for 2 min in 70% (v/v) EtOH, subsequently transferred to a 5% (w/v) NaOCl solution supplemented with Chapter 5 124

0.1 % (v/v) Tween20, and washed with sterile water. the seeds were germinated and grown on half (Arabidopsis) or full strength (tobacco) Murashige and Skoog medium in a growth chamber under a 16-h/8-h light/dark photoperiod at 21°C ± 2°C (Arabidopsis) or 24°C± 2°C (tobacco). The ARR5:GUS lines in cytokinin receptor mutant backgrounds were obtained from the Schmülling lab.

Infection and sampling.

Prior to infections 2 days old R. fascians cultures were washed and concentrated 4 times by resuspending them in sterile water. For tobacco seedling infection seeds were germinated for ± 5 days, until the radicle emerged after which a drop (± 20 µL) of R. fascians culture was applied and phenotypes (growth inhibition, thickening of the vascular tissue) were scored along the interaction or infected seedlings were sampled at desired time points to monitor bacterial fas gene expression. Arabidopsis plants were infected at the developmental stage 1.05 (16 days old with five visible leaves) (Boyes et al., 2001) by applying locally a drop of bacterial culture to the shoot apical meristem and sampled on regular basis for analyzing GUS expression.

In planta determination of fas gene expression.

For the in planta expression analysis tobacco seedlings were infected and sampled as described above. Extracts were prepared by crushing the plants with a pestle in an Eppendorf tube, after which immediately 500 µL of MUG buffer was added (50 mM NaPO4, pH 7,0, 10 mM �-mercapto-ethanol,10 mM EDTA, 0,1% SDS and 0,1% Triton X-100). Ten µL of sample was used for a set of 10-fold dilutions in order to determine the colony forming units (CFU’s). The substrate 4-methylumbelliferyl-�-D-glucoronide was added to a final concentration of 0.1 mM and incubated at 37°C for 1 hour, after which the reaction was stopped by adding a sample of 50 µL to 200 µL of 200 mM Na2CO3 in black nunc polysorb 96-well plates. GUS activity was determined by excitation at 365 nm and measurement at 460 nm (Fluostar optima reader) and is calculated as the measured emission/[time (min) * OD600].

Histochemical staining.

For GUS staining, the entire plant was sampled at 4 and 14 dpi or after 10 days of cytokinin treatment, and subsequently stained and analyzed as described in Depuydt et al. (2008). GUS-marked plants were submerged in 90% (v/v) acetone at 4°C for 1 h and transferred to a

GUS-staining solution of 2 mM 5-bromo-4-chloro-3-indolyl-ß-d-glucuronide and 0.5 mM K3

Fe(CN)6 in buffer containing 100 mM Tris and 50 mM NaCl (pH 7.0). After 19 h of incubation A dedicated role for the Fas proteins in cytokinin production and virulence 125

at 37°C in the dark, the tissue was cleared in 96% EtOH (v/v). Samples were stored in lactic acid (Acros Organics, New Jersey, USA) at 4°C until analysis with a binocular stereomicroscope (Zeiss, Jena, Germany). The three biological repeats each consisted of at least five plants per treatment.

Author contributions. IP gathered data for Figures 2- 4, 6, 8-10, 12; together with KV and LS for Figure 5; with KV for Figure 7; and with SD for Figure 11.

Chapter 6

�

�

�

� De novo cytokinin biosynthesis in � bacteria: an inheritant characteristic � of phytopathogens? �

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

�

� “Bacterial cytokinin production: different pathways for different purposes” � (in preparation). Pertry I, De Backer P, Holsters M and Vereecke D.

De novo cytokinin biosynthesis in bacteria: inheritant to plantpathogens? 129

Introduction

Throughout their growth and development, plants are continuously challenged by biotic agents. Non-pathogenic plant growth-promoting rhizobacteria (PGPRs) stimulate and facilitate plant growth, in part, via phytohormone production (Kloepper and Schroth, 1978; Tsavkelova et al., 2006). Many soil and plant-associated bacteria have been reported to synthesize cytokinins, including PGPRs belonging to the genera Azotobacter, Azospirillum, Arthrobacter, Aerobacter, Acinetobacter, Chromobacterium, Bacillus, Flavobacterium, Micrococcus, Pseudomonas and Rhizobium (Phillips and Torrey, 1972; Barea et al., 1976; Tien et al., 1979; Taller and Wong, 1989; Tsavkelova et al., 2006); methylotrophic and methanotrophic bacteria (Lindstrom and Chistoserdova, 2002; Tsavkelova et al., 2006) and certain Actinomycetes, such as Streptomyces spp. olivaceoviridis, rimosus and rochei; and Frankia sp. (Aldesuquy et al., 1998; Stevens and Berry, 1988). Moreover, the ability to produce cytokinins and disrupt the hormonal balance of the plant, is characteristically related to phytopathogenicity (Jameson, 2000; Tsavkelova et al., 2006) and plays a role in the establishment or severity of a wide array of symptoms. Indeed, the scab causing S. turgidiscabies, blight and rot causing Xanthomonas sp. ; wilt-inducing Ralstonia solanacearum, and hyperplasia-inducing A. tumefaciens, A. rhizogenes, P. savastanoi, P. agglomerans, and R. fascians were all reported to secrete cytokinins (Joshi and Loria, 2007, Barea et al., 1976; Shigaki et al., 2000; Akiyoshi et al., 1987; Lichter et al., 1995a; Chapter 3). Finally, also bacteria that are not associated with plants, but live in the human body, soil or water, such as E. coli, Proteus mirabilis, Bacillus megaterium, B. cereus (Karadeniz et al., 2006), cyanobacteria (Tsavkelova et al., 2006), marine bacteria (Maruyama et al., 1986) and planktonic bacterial genera such as Vibrio, Bacillus, Aeromonas and Achromobacter (Donderski and Gluchowska, 2000) also proved to synthesize cytokinins. With a thorough in silico and phylogenetic analysis, we addressed the questions from where the fas locus could originate; if other bacteria besides S. turgidiscabies possessed (part) of the fas biosynthetic pathway, and if cytokinin biosynthesis via an ipt gene is widespread.

Chapter 6 130

Results and discussion

The fas locus: an island on the linear plasmid

The in silico analysis of the fas locus, described in chapter 4, revealed that the G+C content of the entire fas locus ranged between 53 and 61% (see Table 1), which is much lower than the average 61 to 68% reported for R. fascians (LeChevalier, 1986). This distinct G+C content indicates that the fas locus might be a pathogenicity island (PAI), acquired via horizontal gene transfer (HGT) from a low GC organism. PAIs are characterised by their restricted occurrence in few related species, defining a characteristic virulence phenotype, and by their association with genes related to mobile elements, such as integrases, transposases and insertion sequences, but in many cases the mechanisms mediating gene transfer are uncharacterised (Arnold et al., 2003; Pallen and Wren, 2007). Although the fas locus is not flanked by mobile elements, the entire linear plasmid is conjugative (Crespi et al., 1992). Moreover, a fas orthologous locus is present on a large PAI in S. turgidiscabies, and confers the ability to this scab-causing pathogen to induce leafy galls on tobacco, indistinguishable from those induced by R. fascians (Kers et al., 2005; Joshi and Loria, 2007; see Chapter 4). Unexpectedly, the G+C content of the S. turgidiscabies fas genes was much higher than those in R. fascians, ranging from 56 to 69% (see Table 1). However, the average G+C content of S. turgidiscabies is also higher (71%) than that of R. fascians. The divergent G+C contents of the two fas loci possibly reflects an adaptation of the acquired fas genes to allow efficient expression in both bacterial backgrounds. Indeed, the relative G+C contents of the different fas genes was largely conserved, for instance fasR and fasD had the lowest % G+C in both PAIs. Besides the high similarity of the proteins and the colinear arrangement of the genes, a putative common origin for both fas loci was supported by phylogenetic analyses for mtr1, mtr2, fasD, fasE and fasF (see below).

R. fascians S. turgidiscabies (61-68%) (71%) fasR/araC 53 56 mtr1 57 65 mtr2 59 66 fasA 60 69 fasB 59 66 fasC 60 66 fasD 56 63 fasE 58 66 fasF 61 67

Table 1: G+C content of the genes of the fas locus in R. fascians (Chapter 4) and S. turgidiscabies (Joshi and Loria, 2007). The genomic % G+C range is given between brackets. De novo cytokinin biosynthesis in bacteria: inheritant to plantpathogens? 131

When we analysed the genes surrounding the fas locus to determine the borders of this possible fas PAI, we identified two open reading frames, ORFX (1305 bp) and ORFY (1197 bp), starting 107 bp downstream of fasF, orientated in the same direction (see Figure 1), which exhibited an equally low G+C content of respectively 58 and 59%. The genes belonging to the att operon, immediately upstream of fasR, and the cutinase locus immediately downstream of ORFY (our unpublished results), had a G+C content comparable to that of the genome, thus delimiting the fas PAI, to the 9 genes of the fas locus (see Chapter 4) and ORFX and ORFY.

ORFX ORFY

Figure 1: Schematic representation of the hypothesised fas-associated pathogenicity island.

The association of these two genes with the fas PAI, might imply that their gene products would also be involved in cytokinin production and/or pathogenicity. ORFX and ORFY are separated by 136 bp, suggesting that they make up two separate transcriptional units. To gain insight into their encoded enzymatic function, similarity searches were done with the BLASTP program against the non-redundant protein database at NCBI. ORFX showed striking homology to dibenzothiophene (DBT) desulfurization enzyme A (DszA) of different organisms, ranging from 55% to 69% overall similarity on the protein level (Ishi et al., 2000; Vitorello et al., 2004; Normand et al., 2007). DszA is a mono-oxygenase, part of the sulfur- specific pathway present in bacteria, using DBT as a sole sulfur source (Matsubara et al., 2001). ORFY shows homology to a putative polyketide associated protein from Amycolatopsis mediterranei (51% overall similarity on the protein level) (August et al., 1998) and to acyltransferase polyketide associated protein (Pap) A5 from different organisms (47% to 50% overall similarity on the protein level) (Camus et al., 2002; Ishikawa et al., 2004, Stinear et al., 2007). Although ORFX and ORFY are not conserved in the S. turgidiscabies PAI, arguing against a role in cytokinin biosynthesis, the putative protein functions might point towards an involvement in the de novo biosynthesis of the fas-dependent MeS-type cytokinins (see Chapter 5), by catalyzing the thiolation (ORFX) and methylation (ORFY) of the adenine ring. If this is correct, 2MeS-type cytokinins would be specific for R. fascians.

Chapter 6 132

mtr1, mtr2, fasE and fasF occurrence is not strictly correlated with cytokinin production

The Mtr phylogenetic tree consists of two clusters (see Figure 2). The Mtr1 and Mtr2 proteins of R. fascians and S. turgidiscabies cluster together with putative methyltransferases from different Actinomycetes (see Figure 2), suggesting a common origin. Unfortunately, no functions have been assigned to any of these proteins (Bentley et al., 2002; Ikeda et al., 2003, Choulet et al., 2006; Hayashi et al., 2007; Oliynyk et al., 2007; Ohnishi et al., 2008), Moreover, for none of these organisms cytokinin production has been reported, which might imply that the encoded enzymes have different substrate specificities. The SAM-binding domain of Mtr1 and Mtr2 was also similar to that of sterol- and �-tocopherol methyltransferases from different plants, but these enzymes cluster in the second group, implying different enzymatic functions (see Figure 2). Interestingly, in the same cluster of Mtr1 and Mtr2 are N-methyltransferases from maize and Arabidopsis, suggesting that these enzymes in R. fascians might target a nitrogen atom on the cytokinin molecule.

Figure 2: Phylogenetic analysis of protein homologues of Mtr1 and Mtr2 of R. fascians. Methyltransferases from Streptomyces turgidiscabies (Mtr1: AAW49309.1, Mtr2: AAW49310.1), S. griseus (BAB20508.1), S. ambofaciens (CAJ89345.1), S. argenteolus (BAF98640.1), S. griseus subsp. griseus (YP_001822780.1), Streptomyces sp. Mg1 (ZP_03175267.1), S. coelicolor (NP_631739.1), Mycobacterium marinum M (ref|YP_001849872.1), and Saccharopolyspora erythraea NRRL 2338 (YP_001105918.1). Sterolmethyltransferases from A. thaliana (AAM91592.1) and Zea mays (NP_001106071.1) Gamma-tocopherol methyltransferases from Brassica napus (ACD03287.1), B. oleracea (AAO13806.1), Lotus corniculatus var. Japonicus (AAY52459.1), Triticum aestivum (AAZ67143.1), A. thaliana (AAD02882.1), Gossypium hirsutum (ABE41798.1), Nostoc sp. PCC 7120 (NP_485843.1), Z. mays (NP_001105914.1), Stigmatella aurantiaca DW4/3-1 (ZP_01460522.1), and Nodularia spumigena CCY9414 (ZP_01628101.1). Hypothetical proteins from Ostreococcus lucimarinus CCE9901 (XP_001421685.1), Vitis vinifera (CAO62124.1), Gibberella zeae PH-1 (XP_382959.1), and Oryza sativa (NP_001047844.1). Cyclopropane-fatty-acyl-phospholipid synthase from Anabaena variabilis ATCC 29413 (YP_325298.1) Phosphoethanolamine N-methyltransferase 2 from A. thaliana. (NP_973993.1) and Z. mays (NP_001105267.1). De novo cytokinin biosynthesis in bacteria: inheritant to plantpathogens? 133

Mtr1

Mtr2

Figure 2: Phylogenetic analysis of protein homologues of Mtr1 and Mtr2 of R. fascians.

Chapter 6 134

Within the FasE phylogenetic tree, oxidases and cytokinin oxidases/dehydrogenases group in two subclusters. FasE is part of a the largest subcluster that contains (putative) plant and bacterial CKXs, indicating that these oxidases target cytokinins, possibly with different specificities (see Figure 3). Nevertheless, since no functional data are available for the oxidases from Streptoalloteichus hindustanus (Tao et al., 2007), Myxococcus xanthus (Goldman et al., 2006), Saccharopolyspora erythraea (Oliynyk et al., 2007), Herpetosiphon aurantiacus (YP_001547531.1), Stigmatella aurantiaca (ZP_01464779.1) and Streptomyces pristinaespiralis (YP_002196527.1), their gene function remains hypothetical. Moreover, except for R. fascians and S. turgidiscabies no Ipt gene was predicted in the genomes of these organisms (see Figure 5), so a role in cytokinin metabolism is highly speculative.

Figure 3: Phylogenetic analysis of protein homologues of FasE of R. fascians. Cytokinin oxidases from Legionella pneumophila str. Corby (YP_001251667.1), A. thaliana ATCKX1/CKX1 (NP_181682.1), Streptomyces turgidiscabies (AAW49304.1), Legionella pneumophila subsp. pneumophila str. Philadelphia 1 (YP_094928.1), Stigmatella aurantiaca DW4/3-1 (ZP_01464779.1), Zea mays (NP_001105526.1), and Triticum aestivum (AAK51494.1). Oxidases from Anabaena variabilis ATCC 29413 (YP_325209.1), Herpetosiphon aurantiacus ATCC 23779 (YP_001547531.1), Myxococcus xanthus DK 1622 (YP_634275.1), Streptomyces pristinaespiralis ATCC 25486 (YP_002196527.1), Saccharopolyspora erythraea NRRL 2338 (YP_001108723.1, YP_001103901.1 and YP_001109355.1, Streptomyces avermitilis MA-4680 (NP_826599.1), Mycobacterium tuberculosis H37Rv (NP_216242.1), Polaromonas sp. JS666, Solibacter usitatus Ellin6076 (YP_824467.1), and Streptoalloteichus hindustanus (ABL74933.1). Hypothetical proteins from Nodularia spumigena CCY9414 (ZP_01632120.1), Legionella pneumophila str. Lens (YP_126284.1), Nostoc sp. PCC 7120 (NP_484368.1), and Vitis vinifera (CAO23642.1).

De novo cytokinin biosynthesis in bacteria: inheritant to plantpathogens? 135

Figure 3: Phylogenetic analysis of protein homologues of FasE of R. fascians.

Chapter 6 136

In the FasF phylogenetic tree, there is only a significant distinction between the lysine decarboxylase/phosphoribohydrolases from plants and bacteria. Within the bacterial subcluster, bootstrap values are too low to draw any strong conclusion, so it seems that if a subset of these proteins are involved in cytokinin biosynthesis as phosphoribohydrolases, this specialised activity is not reflected in the primary sequence of the enzymes (see Figure 4). Intriguingly, most bacterial species with a FasF homologue, belong to genera that were reported to produce cytokinins (see Introduction), possibly pointing at an involvement in cytokinin metabolism.

De novo cytokinin biosynthesis mediated by a FasD-like Ipt is correlated to phytopathogenicity

The FasD phylogenetic tree splits in two large clusters: one that contains the tRNA-IPTs and the plant IPTs, and another that has the bacterial Ipts. Within the latter, the Actinomycetes and cyanobacteria cluster separately from the Gram-negatives. Although many bacteria have been reported to produce cytokinins (see Introduction), our database searches on complete bacterial genomes revealed that de novo cytokinin biosynthesis using an Ipt protein seems to be limited to photosynthetic cyanobacteria and plant pathogenic bacteria (see Figure 5).

Figure 4: Phylogenetic analysis of protein homologues of FasF of R. fascians. LOG from Oryzae sativa (). (Lysine) decarboxylases from Bacillus clausii KSM-K16 (YP_175433.1), Geobacter sulfurreducens PCA (NP_953810.1), Legionella pneumophila str. Corby (YP_001251246.1), Legionella pneumophila subsp. pneumophila str. Philadelphia 1 (YP_096506.1), Methylococcus capsulatus str. Bath (YP_114604.1), Ralstonia solanacearum UW551 (ZP_00943328.1), and S. turgidiscabies (AAW49312.1). Hypothetical proteins from A. rhizogenes (NP_066633.1), A. tumefaciens (NP_354090.2), Anaeromyxobacter sp. Fw109-5 (YP_001379786.1), A. thaliana (NP_181258.2), Chromobacterium violaceum ATCC 12472 (NP_902026.1), Pseudomonas fluorescens Pf-5 (YP_257733.1), Pseudomonas putida GB-1 (YP_001671147.1), Ralstonia eutropha JMP134 (YP_295160.1), Rhodospirillum rubrum ATCC 11170 (YP_425449.1), Xanthomonas oryzae pv. oryzae KACC10331 (YP_200680.1), Stigmatella aurantiaca DW4/3-1 (ZP_01466591.1), and Vitis vinifera (CAO18161.1). Nucleotide binding protein from Acinetobacter baumannii ACICU (YP_001844868.1). De novo cytokinin biosynthesis in bacteria: inheritant to plantpathogens? 137

Figure 4: Phylogenetic analysis of protein homologues of FasF of R. fascians. Species indicated with belong to genera reported to produce cytokinins (see Introduction).

Chapter 6 138

Cytokinins can originate from two pathways: newly synthesised via an Ipt or released via tRNA degradation (see Chapter 1 and references therein). Interestingly, also for bacterial auxin biosynthesis, two different pathways exist: the indole-3-pyruvate pathway (iPyA) and the indole-3-acetamide (IAM) pathway. The former is involved in (epiphytic) fitness and occurs both in pathogenic and non-pathogenic bacteria, whereas the latter is correlated with virulence and rhizobial symbiosis (Spaepen et al., 2007 and references therein). By analogy, it is tempting to assign a similar differential role for both cytokinin pathways: tRNA-mediated biosynthesis would be a very general pathway, whereas the Ipt pathway would be required for pathological activities. Generally, nodule formation is induced by bacterial Nod-factors (Mergaert et al., 1997). However, Nod- mutants overproducing cytokinins and natural Bradyrhizobium strains lacking nod genes, also induce nodule formation (Cooper and Long, 1994; Giraud et al., 2007). Moreover, cytokinin perception mutants of Medicago truncatula are no longer capable of forming nodules (Murray et al., 2007), illustrating the important role of cytokinins in nodulation. Interestingly, Bradyrhizobium japonicum and Rhizobium leguminosarum have been reported to produce cytokinins (Phillips and Torrey, 1972; Sturtevant and Taller, 1989), yet no FasD-homologues were identified in their genomes (Young et al., 2006; Giraud et al., 2007; see Figure 5), suggesting that cytokinins in these organisms originate from tRNA degradation and supporting our hypothesis.

Figure 5: Phylogenetic analysis of protein homologues of FasD of R. fascians. Ipts from A. tumefaciens ( ; P15653.1), A. vitis (AAB41870.1), Anabaena variabilis ATCC 29413 (YP_323219.1), Dictyostelium discoideum AX4 (XP_642693.1), Nostoc sp. PCC 7120 (NP_484264.1), Ralstonia solanacearum (#; NP_522786.1), S. turgidiscabies (AAW49305.1), Xanthomonas oryzae pv. oryzae KACC10331 (YP_201575.1), Z. mays (ABY78887.1), Lotus japonicus (ABD93932.1), and A. thaliana (AtIPT4)(NP_194196.1). Tzs (tZ producing) from A. rhizogenes (P14011.1), A. tumefaciens ( ; NP_396682.2 ), P. savastanoi (P06619.1), and R. solanacearum (§; P14333.1). tRNA Ipts from E. coli (AP_004671.1 and YP_002331946.1), Clavibacter michiganensis subsp. michiganensis NCPPB 382 (CAN02085.1), Leifsonia xyli subsp. xyli str. CTCB07, Rhodococcus sp. RHA1 (YP_706705.1), A. thaliana (AtIPT2) (NP_565658.1), S. cerevisiae (P07884.2), and A. tumefaciens ( ; NP_355007.2). Hypothetical proteins from Vitis vinifera (CAN73470.1) and Oryza sativa (EAY98937.1).

De novo cytokinin biosynthesis in bacteria: inheritant to plantpathogens? 139

§ #

Figure 5: Phylogenetic analysis of protein homologues of FasD of R. fascians.

Chapter 6 140

Concluding remarks The origin of the fas PAIs of R. fascians and S. turgidiscabies remains speculative. Horizontal gene transfer is mainly reported between prokaryotes, nevertheless transfer from eukaryotes to prokaryotes has been reported and may play a role in bacterial pathogenicity (Koonin et al., 2001; Pallen and Wren, 2007). The similarity of different fas proteins to plant homologues, suggests that the fas genes might indeed be plant-derived. Moreover, we postulate that Ipt-derived cytokinins are correlated with virulence, whereas tRNA-derived cytokinins are a general bacterial feature. Despite the central position of Ipt proteins in cytokinin biosynthesis, our phylogenetic analysis shows that the R. fascians and S. turgidiscabies Ipts are different from those of the Gram-negative plant pathogens, possibly indicating distinctive substrate specificities and eventually reflecting specific pathologies. Finally, the fas-mediated cytokinin biosynthetic machinery is uniquely occurring in the two leafy gall-inducing bacteria known to date, R. fascians and S. turgidiscabies, implying that it is responsible for the production of a specialised set of molecules that manipulate plant development in a very characteristic way.

Material and Methods In silico analysis. The program ORF Finder (NCBI) was used to detect the presence of open reading frames, which were translated to their corresponding amino acid sequences by using the Expasy translation tool. Homologous proteins were identified with the Basic Local Alignment Tool (BLAST) against the non-redundant protein database at NCBI. The percentage GC was determined via the GC calculator program (http://www.genomicsplace.com/gc_calc.html).

Phylogenetic analysis. The Maximum-likelihood tree was based on a selection of genes based on E-value scores after Blastp, starting from the relevant gene from the fas locus of R. fascians D188. Sequence data and annotation files were available and downloaded from the NCBI Microbial Genome Resource database (The NCBI Microbial Genome Resource Database [http://www.ncbi.nlm.nih.gov/genomes/MICROBES/microbial_taxtree.html]). Sequences were aligned using the Clustalw program (Thompson, 2002). The phylogenetic tree was constructed with the Phyml program (Guindon, 2003). A WAG substitution model (Whelan, 2001) and 100 bootstrap replicates were run. Unless indicated otherwise, bootstraps are 100.

Author contributions. IP gathered data for Table 1 and Figure 1; and together with PDB for Figures 2-5.

Summary and perspectives

Summary and perspectives 143

Phytohormones play a dual role in plants: besides controlling plant growth and development, they also determine the outcome of plant-microbe interactions. Plant pathogens, such as Agrobacterium tumefaciens, A. rhizogenes, Pseudomonas savastanoi, and Pantoea agglomerans, have evolved to disrupt the plants defined hormonal balance either by genetically transforming their host or by producing these phytohormones themselves. Besides auxins, cytokinins also are virulence determinants in these strains, and facilitate or establish disease development. For a long time tRNA was thought to be the source of cytokinins. However, genes were identified in these phytopathogens, which encoded the enzymatic machinery for de novo cytokinin biosynthesis, eventually resulting in the characterisation of in planta cytokinin metabolism (Chapter 1). Amongst these gall-inducing bacteria, the Actinomycete Rhodococcus fascians has a somewhat unique position, since it induces leaf deformations, shooty outgrowths and differentiated leafy galls. The only other known phytopathogen capable of doing so, is Streptomyces turgidiscabies (Chapters 1 and 2). Just like other hyperplasia inducers, R. fascians produces both cytokinins and auxins. Auxin production was hypothesized to be involved in epiphytic fitness and full symptom development. However, the role of bacterial auxin production in symptomatology remains to be elucidated (Chapter 2). The link between R. fascians cytokinin production, and virulence has been debated since 1966. A significant breakthrough concerning this issue was the identification of an isopentenyltransferase on a linear virulence plasmid of R. fascians strain D188, which was essential for and strictly correlated with virulence. The results of the presented work underline the central role of cytokinins in disease establishment by demonstrating that cytokinin perception by the host is essential for virulence. We showed that R. fascians produces a mix of 6 types of cytokinins (iP, cZ, tZ and their 2MeS-derivatives), which have strong synergistic effects and are perceived in Arabidopsis via the cytokinin receptors AHK3 and AHK4, which is upregulated upon infection. Simultaneous disruption of both receptors resulted in the abolishment of symptoms. Unexpectedly, the linear-plasmid free strain produced the same array of cytokinins, indicating that R. fascians virulence is a matter of cytokinin ratios rather than the production of specialised molecules. We postulate that the continuous challenge of the host with a mixture of synergistically acting cytokinins, lies at the basis of R. fascians pathology. The plant homeostasis mechanisms, activated in an attempt to counter the disrupted hormonal balances, are inadequate. The CKX enzymes do not efficiently degrade cis-type cytokinins resulting in their accumulation and the maintenance of symptoms and tissue proliferation throughout the interaction (Chapter 3). The only plausible candidate on the linear plasmid for cytokinin production was the fas locus, and we showed that it indeed encodes the cytokinin machinery. The re-evaluation of the sequence and analyses of the cytokinin profiles and the virulence phenotypes of the Summary and perspectives 144

different fas mutants, indicated how the fas locus contributes to R. fascians cytokinin production and unambiguously identified cZ and tZ as the main fas cytokinins, since their production required the entire fas locus (Chapters 4 and 5). Altogether, these data resulted in a model where each Fas protein had a dedicated role in cytokinin production and symptom development. The key enzyme for virulence and virulence-associated cytokinin production is FasD, which synthesizes the precursor iP. A central role was also assigned to the production of zeatin-type cytokinins, mediated by FasA, which is similarly required for virulence. FasF would direct the alternative production of tZ, 2MeStZ, cZ via its phosphoribohydrolase activity and although this is not essential for symptom induction, it is absolutely required for symptom maintenance. The role of fasE in the determination of the cytokinin spectrum and virulence is more difficult to interprete. However, the available data point to an involvement in controlling cytokinin ratios. Although, both Mtrs influence the production of the six cytokinins and are essential for virulence, their exact biosynthetic function remains to be elucidated.

ADP (/AMP/ATP) + DMAPP Chromosome FasD FasE adenine + 3-methyl-2-butenal iPRMP 2MeSiP iP FasA FasA-C FasF FasA + FasF + Mtr1 & 2 2MeScZ

tZ, 2MeStZ and cZ

Figure 1: Working model for Fas-mediated cytokinin production in R. fascians D188.

The virulence phenotypes of the fas mutants suggest that symptom initiation and maintenance is mediated by a quantitatively and qualitatively controlled release of a cytokinin spectrum, which is underlined by the complex and tightly controlled regulation of fas gene expression. The in vitro and in planta kinetics of fas gene expression imply that the different regulatory mechanisms control the switch of producing high levels of fas-cytokinins to iniate symptoms to a much lower level or a different mix, sufficient for symptom maintenance. This hypothesis is further supported by the transient activation of AHK4-mediated cytokinin signal transduction (Chapters 4 and 5). Summary and perspectives 145

Phylogenetic analysis indicated that although de novo cytokinin production in bacteria is a trait of plant pathogens, R. fascians has evolved a distinct cytokinin biosynthetic pathway, which is only conserved in S. turgidiscabies, indicating that its “trick with the cytokinin mix” is a quite unique concept in phytopathology (Chapter 6).

Although we did identify the virulence-associated cytokinins produced by the fas operon, and cytokinin perception has proven to be essential for symptomatology, we could not mimic the phenotypes completely by treating plants with their mixtures. This might partially be explained by the fact that we can not copy the local and continuous delivery of bacterial cytokinins throughout the interaction. On the other hand, we can not exclude that other virulence effectors are at play. Although a function could be assigned to most Fas proteins, the exact role of Mtr1 and Mtr2, FasE, ORFX and ORFY remains to be elucidated. We can not rule out that they are involved in the production of a virulence factor that awaits discovery. With the collection of fas mutants generated during this study, we could initiate a broader metabolic profiling, which might clarify these points. Most of the plant-pathogens described in Chapter 1 use auxins aside of cytokinin to induce and establish galls on their hosts. In this aspect, the linear-plasmid controlled auxin chromosomal biosynthesis in R. fascians might be very important. To address the role of auxins in R. fascians pathology, the auxin biosynthetic genes will have to be identified, truncated and the infection phenotype of the resulting strains characterized. Meanwhile, different cytokinin/auxin ratios and mixtures could already be tested for their capacity to phenocopy symptoms. Based on the cytokinin profiles, we proposed a de novo biosynthesis for 2MeS- derivatives. However, the genes responsible for this modification still have to be identified. Two possible candidates located immediately downstream of the fas operon seem to be part of the fas pathogenicity island (Chapter 6). Mutants in these genes will have to be generated, and their cytokinin profiles and capacity to induce/maintain virulence will have to be analyzed in detail. The mechanisms regulating fas gene expression are very complex, and although this work contributed to the elucidation of these regulatory pathways, many issues remain to be resolved. To date, the translational regulator has not yet been identified. The hypothesis proposed in Chapter 4, will have to be thoroughly addressed. To assess the role of A-factors in fas regulation, a mutant will have to be generated, incapable of producing these signal molecules, and its virulence and the effect on fas gene expression analysed .

In conclusion, our data have largely uncovered the role of cytokinins and the fas locus in the R. fascians pathology: the continuous challenge with defined ratios of synergistically acting Summary and perspectives 146

cytokinins eventually defeats nearly all plants and transforms them into shooty niches. Many intriguing questions derived from the novel insights obtained during this work remain to be answered. Nevertheless, we feel that the results presented here have shed some light on the remaining secrets of this fascinating pathogen.

Nederlandstalige samenvatting

Nederlandstalige samenvatting 149

Phytohormonen kunnen een tweezijdige rol spelen in planten: naast het reguleren van plantengroei en –ontwikkeling, spelen ze ook een belangrijke rol in plant-bacterie interacties. Fytopathogenen, zoals Agrobacterium tumefaciens, A. rhizogenes, Pseudomonas savastanoi, en Pantoea agglomerans kunnen de strikt gereguleerde hormoon balansen van de plant verstoren door de plant genetisch te transformeren om deze hormonen the produceren of door deze hormonen zelf te produceren en secreteren. In deze stammen vormen cytokinines, naast auxines, belangrijke virulentiefactoren die symptomen induceren of de ontwikkeling ervan beïnvloeden. Gedurende lange tijd werden cytokinines beschouwd als tRNA afbraakproducten. De novo cytokinine biosynthese werd voor het eerst opgehelderd in deze organismen met de identificatie van de genen en enzymen die deze reacties katalyseren, nl. isopentenyltransferasen. Hun gedetailleerde studie resulteerde uiteindelijk eveneens in het identificeren en ophelderen van cytokinine biosynthese in planta Cytokinine perceptie en signalisatie reguleert de genexpressie van een wijde variëteit aan genen, die betrokken zijn bij scheut- en wortelgroei, vascularisatie, fertiliteit, nutriënt mobilisatie, stimulatie van celdeling, senescentie, … (Hoofdstuk 1). De Actinomyceet R. fascians heeft een redelijk unieke positie onder deze hyperplasia inducerende bacteriën. Deze bacterie heeft een zeer breed gastheerbereik en induceert naast bladmisvormingen, de vorming van gedifferentieerde gallen op zijn gastheer die bestaan uit scheuten, geïnhibeerd in hun uitgroei. Deze gallen vormen specifieke niches voor de bacteriën en voorzien hen van allerlei nutriënten. R. fascians heeft een lineair virulentieplasmide waarop drie virulentie loci geïdentificeerd zijn: fas, att en hyp. De genen van het att locus coderen voor eiwitten die verantwoordelijk zijn voor de productie van een autoregulatorische component die onder andere betrokken is bij regulatie van fas gen expressie. Terwijl mutaties in het att locus verantwoordelijk zijn voor de ontwikkeling van minder of kleinere bladgallen, resulteren hyp mutaties in een hypervirulent fenotype. Het fas locus vormt het onderwerp van deze thesis en is essentieel voor virulentie. Het codeert voor eiwitten die homoloog zijn met de enzymen verantwoordelijk voor cytokinine biosynthese. De strikte regulatie van dit operon, bekrachtigt het belang ervan voor virulentie. Bovendien werd het volledige fas operon geïdentificeerd in Streptomyces turgidiscabies, de enige andere plantpathogeen die de vorming van gedifferentieerde gallen kan induceren op zijn gastheer (Hoofdstuk 2). Aangezien de aard van de symptomen aangeeft dat cytokinines belangrijke virulentie- factoren zijn voor R. fascians, werd reeds sinds 1966 getracht hun link met R. fascians virulentie op te helderen. R. fascians produceert niet minder dan 11 cytokinines maar tegenstrijdige data in planta maakten het onmogelijk om deze met pathogeniciteit te linken. Meer duidelijkheid werd gebracht met de identificatie van een isopentenyltransferase gen op het lineair plasmide (fasD), essentieel en strikt gecorreleerd met virulentie en bovendien tot Nederlandstalige samenvatting 150

expressie gebracht onder zeer specifieke omstandigheden die de aanwezigheid van de gastheerplant weerspiegelen. De resultaten bekomen in dit werk bevestigen de centrale rol van cytokinines in R. fascians virulentie, gezien cytokinine perceptie door de gastheer essentieel bleek te zijn voor de ontwikkeling van symptomen. Het lineair plasmide draagt bij tot de productie van zes cytokinines (iP, 2MeSiP, cZ, 2MeScZ, tZ en 2MeStZ) die continu gesecreteerd worden en een sterk synergistisch effect hebben. In een poging deze verstoring van de hormoonbalans te counteren, activeert de plant de expressie van cytokinine dehydrogenasen die de afbraak van cytokinines bewerkstelligen. In Arabidopsis worden cis-type cytokinines echter niet efficiënt afgebroken, wat resulteert in hun accumulatie en het onderhouden van weefselproliferatie en de symptomen. Gebaseerd op de resultaten, stellen we dat de productie van een mix aan cytokinines R. fascians in staat stelt om het defensiemechanisme van een groot aantal gastheren te overwinnen: naargelang de gastheer zullen andere, voor afbraak resistente, cytokinines accumuleren en symptoom- ontwikkeling in stand houden (Hoofdstuk 3). De enige kandidaat op het lineair plasmide voor het bijdragen tot deze cytokinine biosynthese is het fas locus. Het opnieuw evalueren van de sequentie van dit locus en het bepalen van de cytokinine profielen en virulentiecapaciteit van verschillende fas mutanten, heeft aangetoond hoe dit locus bijdraagt tot cytokinine productie en leidde tot de identificatie van cZ, tZ en 2MeStZ als de fas cytokinines (Hoofdstukken 4 en 5). De resultaten bekomen in dit werk leiden tot een model waarbij elk fas eiwit een specifieke rol heeft in cytokinine biosynthese en symptoomontwikkeling. FasD vormt het centrale eiwit via de productie van iP, dat als precursor dient voor de andere eiwitten. FasA bewerkstelligt de vorming van zeatines door het hydroxyleren van iP of het door het chromosoom geproduceerde 2MeSiP. FasF zou bijdragen tot een alternatieve biosynthese, essentieel voor behoud van symptomen, van cZ, tZ en 2MeStZ door deze cytokinines direct vrij te stellen uit hun nucleotide vormen. Hoewel Mtr1, Mtr2 en FasE duidelijk betrokken zijn bij cytokinine biosynthese, dienen hun eigenlijke rol in dit proces en in virulentie nog opgehelderd te worden. FasE codeert voor een cytokinine dehydrogenase en zou dus mogelijks de vrijgestelde cytokinine verhoudingen kunnen controleren. De fenotypes van de verschillende fas mutanten geven aan dat symptoomontwikkeling het gevolg is van een strikt kwalitatief en kwantitatief gereguleerde vrijstelling van een cytokinine spectrum. Dit wordt gereflecteerd door de strikte regulatie van het fas operon, de differentiële expressie kinetiek in vitro en in planta en de transiënte activatie van cytokinine signaal transductie gemedieerd door de cytokinine receptor AHK4. Fylogenetische analyses gaven aan dat, hoewel de novo cytokinine biosynthese in bacteriën kenmerkend zou zijn voor plantpathogenen, de cytokinine biosynthese pathway Nederlandstalige samenvatting 151

zoals die in R. fascians verder enkel behouden is in S. turgidiscabies (Hoofdstuk 6), wat het uniek karakter aantoont van deze mix strategie.

ADP (/AMP/ATP) + DMAPP Chromosome FasD FasE adenine + 3-methyl-2-butenal iPRMP 2MeSiP iP FasA FasA-C FasF FasA + FasF + Mtr1 & 2 2MeScZ

tZ, 2MeStZ and cZ

Figuur 1: Model voor Fas-gemedieerde cytokinine productie in R. fascians D188.

Hoewel we erin geslaagd zijn de door het fas operon en met virulentie geassocieerde cytokinines te identificeren en aangetoond hebben dat cytokinine perceptie door de plant essentieel is voor symptoomontwikkeling, was het niet mogelijk om de symptomen op de plant volledig na te bootsen door ze enkel en alleen met een mengeling van deze cytokinines te behandelen. Dit kan deels verklaard worden doordat het niet mogelijk was om de lokale en continue toevoer van bacteriële cytokinine mixen doorheen de interactie na te bootsen. Anderzijds kunnen we niet uitsluiten dat nog andere virulentiefactoren een rol spelen. Hoewel aan de meeste Fas eiwitten een rol kon toegewezen worden, dient de exacte functie van Mtr1, Mtr2, FasE, ORFX and ORFY nog bepaald te worden. Het is niet onmogelijk dat deze eiwitten betrokken zijn bij de productie van een nog te identificeren virulentiefactor. Deze vragen zouden kunnen opgehelderd worden door een metabolische studie uit te voeren met de verschillende fas mutanten die we tijdens dit werk gecreëerd hebben. De meeste plantpathogene bacteriën beschreven in Hoofdstuk 1 gebruiken naast cytokinines ook auxines om hun symptomen te bewerkstelligen. Vanuit dit inzicht, is het feit dat R. fascians ook auxines produceert, vanuit het chromosoom maar wel lineair plasmide afhankelijk, een interessant en mogelijks belangrijk gegeven. Om de rol van auxines in de pathologie van R. fascians te bepalen, dienen de auxine biosynthese genen geïdentificeerd en gemuteerd en hun effect op virulentie gekarakteriseerd te worden. In afwachting kunnen verschillende cytokinine/auxine verhoudingen en mixen getest worden op hun capaciteit om de door R. fascians veroorzaakte fenotypes volledig na te bootsen. Nederlandstalige samenvatting 152

Op basis van de cytokinine profielen, hebben we vooropgesteld dat 2MeS-derivaten volledig nieuw geproduceerd worden. De genen verantwoordelijk voor deze modificatie dienen evenwel nog geïdentificeerd te worden. Onmiddellijk na het fas operon bevinden zich twee mogelijke kandidaten die waarschijnlijk ook deel uitmaken van het fas pathogeniciteitseiland (Hoofdstuk 6). Mutanten in deze genen zullen gecreëerd moeten worden en hun cytokinine profielen en pathologie in detail geanalyseerd worden. Het fas gen expressiemechanisme is zeer complex en hoewel dit werk heeft bijgedragen tot de verdere opheldering van deze pathways, blijven veel vragen nog onbeantwoord. Tot op vandaag blijft de translationele regulator nog ongekend. De hypothese vooropgesteld in Hoofdstuk 4 over de betrokkenheid van A-factoren, zal grondig getest moeten worden door mutanten te generen die niet langer in staat zijn deze signaalmoleculen te produceren en hun virulentie en effect op expressie na te gaan.

Tot besluit kunnen we stellen dat onze data een groot deel hebben opgehelderd over de rol van cytokinines en het fas locus in de pathologie van R. fascians: de voortdurende blootstelling aan door de bacterie strikt gedefinieerde verhoudingen van synergistisch werkende cytokinines zal uiteindelijk bijna elke plant overwinnen en transformeren tot specifieke niches. Veel intrigerende vragen vloeien voort uit de nieuwe inzichten die we met dit werk bereikt hebben. En hoewel deze nog dienen opgehelderd te worden, vinden we dat de gepresenteerde data bijgedragen hebben tot het ontrafelen van de geheimen van deze fascinerende bacterie.

References

References 155

Abe I, Tanaka H, Abe T and Noguchi H (2007) Enzymatic formation of unnatural cytokinin analogs by adenylate isopentenyltransferase from mulberry. Biochem Biophys Res Commun. 355(3):795-800. Akiyoshi DE, Morris RO, Hinz R, Mischke BS, Kosuge T, Garfinkel DJ, Gordon MP and Nester EW (1983) Cytokinin/auxin balance in crown gall tumors is regulated by specific loci in the T-DNA. Proc. Natl. Acad. Sci. USA 80: 407-411. Akiyoshi DE, Klee H, Amasino RM, Nester EW and Gordon MP (1984) T-DNA of Agrobacterium tumefaciens encodes an enzyme of cytokinin biosynthesis. Proc. Natl. Acad. Sci. USA 81:5994- 5998. Akiyoshi DE, Regier DA, Jen G and Gordon MP (1985) Cloning and nucleotide sequence of the tzs gene from Agrobacterium tumefaciens strain T37. Nucl. Acids Res. 13:2773-2788. Akiyoshi DE, Regier DA and Gordon MP (1987) Cytokinin production by Agrobacterium and Pseudomonas spp. J. Bacteriol. 169:4242-4248. Aldesuquy HS, Mansour FM and Abo-Hamed SA (1998) Effect of the culture filtrates of Streptomyces on growth and productivity of wheat plants. Folia Microbiologica 43(5):465-470. Allen M, Qin W, Moreau F and Moffatt B (2002) Adenine phosphoribosyltransferase isoforms of Arabidopsis and their potential contributions to adenine and cytokinin metabolism. Physiol Plant. 115(1):56-68. Anton BP, Saleh L, Benner JS, Raleigh EA, Kasif S and Roberts RJ (2008) RimO, a MiaB-like enzyme, methylthiolates the universally conserved Asp88 residue of ribosomal protein S12 in Escherichia coli. Proc Natl Acad Sci U S A 105(6):1826-31. Appleby CA and Morton RK (1959) Lactic dehydrogenase and cytochrome b2 of baker's yeast. Biochem J 71:492-499. Argyros RD, Mathews DE, Chiang YH, Palmer CM, Thibault DM, Etheridge N, Argyros DA, Mason MG, Kieber JJ and Schaller GE (2008) Type B response regulators of Arabidopsis play key roles in cytokinin signaling and plant development. Plant Cell 20(8):2102-16. Armstrong DJ and Firtel RA (1989) Cytokinin oxidase activity in the cellular slime mold, Dictyostelium discoideum. Dev. Biol. 136(2):491-9. Armstrong DJ, Scarbrough E, Skoog F, Cole DL and Leonard NJ (1976) Cytokinins in Corynebacterium fascians cultures. Isolation and identification of 6-(4-hydroxy-3methyl-cis-2- butenylamino)-2-methylthiopurine. Plant Physiol. 58:749-752. Arnold DL, Pitman A and Jackson RW (2003) Pathogenicity and other genomic islands in plant pathogenic bacteria. Mol. Plant Pathol. 4(5):407-420. Astot C, Dolezal K, Nordström A, Wang Q, Kunkel T, Moritz T, Chua NH and Sandberg G (2000) An alternative cytokinin biosynthesis pathway. Proc Natl Acad Sci U S A 97(26):14778-83. August PR, Tang L, Yoon YJ, Ning S, Mueller R, Yu TW, Taylor M, Hoffmann D, Kim CG, Zhang X, Hutchinson CR and Floss HG (1998) Biosynthesis of the ansamycin antibiotic rifamycin: deductions from the molecular analysis of the rif biosynthetic gene cluster of Amycolatopsis mediterranei S699. Chem. Biol. 5(2):69-79. Balázs E and Sziráki I (1974) Altered levels of indoleacetic acid and cytokinin in geranium stems infected with Corynebacterium fascians. Acta Phytopathol. Acad. Sci. Hung. 9:287-292. Barash I and Manulis-Sasson S (2007) Virulence mechanisms and host specificity of gall-forming Pantoea agglomerans. Trends Microbiol. 15(12):538-545. Barea JM, Navarro E and Montoya E (1976) Production of Plant Growth Regulators by Rhizosphere Phosphate-solubilizing Bacteria. J. Appl. Bact. 40:129-134. Barry GF, Rogers SG, Fraley RT and Brand L (1984) Identification of a cloned cytokinin biosynthetic gene. Proc. Natl. Acad. Sci. USA 81:4776-4780. Bassil NV, Mok D and Mok MC (1993) Partial Purification of a cis-trans-Isomerase of Zeatin from Immature Seed of Phaseolus vulgaris L. Plant Physiol. 102(3):867-872. References 156

Bell KS, Philip JC, Aw DWJ and Christofi N (1998) A review: The genus Rhodococcus. Journal of Applied Microbiology 85:195-210. Bentley SD, Chater KF, Cerdeno-Tarraga AM, Challis GL, Thomson NR, James KD, Harris DE, Quail MA, Kieser H, Harper D, Bateman A, Brown S, Chandra G, Chen CW, Collins M, Cronin A, Fraser A, Goble A, Hidalgo J, Hornsby T, Howarth S, Huang CH, Kieser T, Larke L, Murphy L, Oliver K, O'Neil S, Rabbinowitsch E, Rajandream MA, Rutherford K, Rutter S, Seeger K, Saunders D, Sharp S, Squares R, Squares S, Taylor K, Warren T, Wietzorrek A, Woodward J, Barrell BG, Parkhill J and Hopwood DA (2002) Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417 (6885), 141-147. Bilyeu KD, Cole JL, Laskey JG, Riekhof WR, Esparza TJ, Kramer MD and Morris RO (2001) Molecular and biochemical characterization of a cytokinin oxidase from maize. Plant Physiol. 125(1):378-86. Blackwell JR and Horgan R (1994) Cytokinin biosynthesis by extracts of Zea mays. Phytochemistry 35(2):339-342. Boyes DC, Zayed AM, Ascenzi R, McCaskill AJ, Hoffman NE, Davis KR and Görlach J (2001) Growth stage-based phenotypic analysis of Arabidopsis: A model for high throughput functional genomics in plants. Plant Cell 13:1499-1510. Brown NA (1927). Sweet pea fasciation, a form of crowngall. Phytopathology 17:29-30. Brugière N, Humbert S, Rizzo N, Bohn J and Habben JE (2008) A member of the maize isopentenyl transferase gene family, Zea mays isopentenyl transferase 2 (ZmIPT2), encodes a cytokinin biosynthetic enzyme expressed during kernel development. Cytokinin biosynthesis in maize. Plant Mol Biol. 67(3):215-29. Brzobohatý B, Moore I, Kristoffersen P, Bako L, Campos N, Schell J, Palme K (1993) Release of active cytokinin by a beta-glucosidase localized to the maize root meristem. Science 262(5136):1051-4. Bulgakov VP (2008) Functions of rol genes in plant secondary metabolism. Biotechnology Advances 26:318–324. Bürkle L, Cedzich A, Döpke C, Stransky H, Okumoto S, Gillissen B, Kühn C and Frommer WB (2003) Transport of cytokinins mediated by purine transporters of the PUP family expressed in phloem, hydathodes, and pollen of Arabidopsis. Plant J. 34(1):13-26. Burr TJ, Otten L (1999) CROWN GALL OF GRAPE: Biology and Disease Management. Annu Rev Phytopathol. 37:53-80. Caillet J and Droogmans L (1988) Molecular cloning of the Escherichia coli miaA gene involved in the formation of delta 2-isopentenyl adenosine in tRNA. J Bacteriol. 170(9):4147-52. Camus JC, Pryor MJ, Medigue C and Cole ST (2002) Re-annotation of the genome sequence of Mycobacterium tuberculosis H37Rv. Microbiology (Reading, Engl.) 148(Pt 10):2967-2973. Casanova E, Trillas MI, Moysset L and Vanstein A (2005) Influence of rol genes in floriculture. Biotechnology Advances 23:3 –39. Chalupowicz L, Barash I, Schwartz M, Aloni R and Manulis S (2006) Comparative anatomy of gall development on Gypsophila paniculata induced by bacteria with different mechanisms of pathogenicity. Planta 224(2):429-37. Chatfield JM and Armstrong DJ (1986) Regulation of cytokinin oxidase activity in callus tissues of Phaseolus vulgaris L. cv. Great Northern. Plant Physiology 80:493-499. Chatfield SP, Stirnberg P, Forde BG and Leyser O (2000) The hormonal regulation of axillary bud growth in Arabidopsis. Plant journal 24(2):159-169. Chatterjee A, Cui Y, Liu Y, Dumenyo CK and Chatterjee AK (1995) Inactivation of rsmA leads to overproduction of extracellular pectinases, cellulases, and proteases in Erwinia carotovora subsp. carotovora in the absence of the starvation/cell density-sensing signal, N-(3-oxohexanoyl)-L- homoserine lactone. Appl Environ Microbiol. 61(5):1959-67. Che P, Gingerich DJ, Lall S and Howell SH (2002) Global and hormone-induced gene expression changes during shoot development in Arabidopsis.Plant Cell. 14(11):2771-85. References 157

Chen CM and Melitz DK (1979) Cytokinin biosynthesis in a cell-free system from cytokinin-autotrophic tobacco tissue cultures. FEBS Lett. 107(1):15-20. Chen CM and Kristopeit SM (1981a) Metabolism of Cytokinin : DEPHOSPHORYLATION OF CYTOKININ RIBONUCLEOTIDE BY 5'-NUCLEOTIDASES FROM WHEAT GERM CYTOSOL. Plant Physiol. 67(3):494-498. Chen CM and Kristopeit SM (1981b) Metabolism of Cytokinin: Deribosylation of Cytokinin Ribonucleoside by Adenosine Nucleosidase from Wheat Germ Cells. Plant Physiol. 68(5):1020- 1023. Cherayil JD and Lipsett MN (1977) Zeatin ribonucleosides in the transfer ribonucleic acid of Rhizobium leguminosarum, Agrobacterium tumefaciens, Corynebacterium fascians, and Erwinia amylovora. J Bacteriol. 131(3):741-4. Chory J, Reinecke D, Sim S, Washburn T and Brenner M (1994) A role for cytokinins in de-etiolation in Arabidopsis. det mutants have an altered response to cytokinins. Plant Physiol 104:339-347. Choulet F, Aigle B, Gallois A, Mangenot S, Gerbaud C, Truong C, Francou FX, Fourrier C, Guerineau M, Decaris B, Barbe V, Pernodet JL and Leblond P (2006) Evolution of the terminal regions of the Streptomyces linear chromosome. Mol. Biol. Evol. 23(12):2361-2369. Chung KM, Igari K, Uchida N and Tasaka M (2008) New perspectives on plant defense responses through modulation of developmental pathways. Mol Cells. 26(2):107-12. Cooper JB and Long SR (1994) Morphogenetic Rescue of Rhizobium meliloti Nodulation Mutants by trans-Zeatin Secretion. Plant Cell. 6(2):215-225. Coppins RL, Hall KB and Groisman EA (2007) The intricate world of riboswitches. Curr Opin Microbiol. 10(2):176-81. Cornelis K, Ritsema T, Nijsse J, Holsters M, Goethals K and El Jaziri M (2001) The plant pathogen Rhodococcus fascians colonizes the exterior and interior of aerial parts of plants. Mol. Plant-Microbe Interact. 14:599-608. Cornelis K, Maes T, Jaziri M, Holsters M and Goethals K (2002) Virulence genes of the phytopathogen Rhodococcus fascians show specific spatial and temporal expression patterns during plant infection. Mol. Plant-Microbe Interact. 15:398-403. Cregg JM, Barringer KJ and Hessler AY (1985) Pichia pastoris as a host system for transformations. Mol Cell Biol 5:3376-3385. Crespi M, Messens E, Caplan AB, van Montagu M and Desomer J (1992) Fasciation induction by the phytopathogen Rhodococcus fascians depends upon a linear plasmid encoding a cytokinin synthase gene. EMBO J. 11(3):795-804. Crespi M, Vereecke D, Temmerman W, Van Montagu M and Desomer J (1994) The fas operon of Rhodococcus fascians encodes new genes required for efficient fasciation of host plants. J Bacteriol 176:2492-2501. Cserzo M, Wallin E, Simon I, von Heijne G and Elofsson A (1997) Prediction of transmembrane alpha- helices in procariotic membrane proteins: the Dense Alignment Surface method. Prot. Eng. 10(6): 673-676. Cui Y, Madi L, Mukherjee A, Dumenyo CK and Chatterjee AK (1996) Mol Plant Microbe Interact. 9(7):565-73. Cui Y, Chatterjee A, Yang H and Chatterjee AK (2008) Regulatory network controlling extracellular proteins in Erwinia carotovora subsp. carotovora: FlhDC, the master regulator of flagellar genes, activates rsmB regulatory RNA production by affecting gacA and hexA (lrhA) expression.J Bacteriol. 190(13):4610-23. De Cleene M and De Ley J (1976) The host range of crown gall. The botanical review 42(4):389-413. Dello Ioio R, Linhares FS, Scacchi E, Casamitjana-Martinez E, Heidstra R, Costantino P and Sabatini S (2007) Cytokinins determine Arabidopsis root-meristem size by controlling cell differentiation. Curr Biol. 17(8):678-82. References 158

Dello Ioio R, Linhares FS and Sabatini S (2008) Emerging role of cytokinin as a regulator of cellular differentiation. Curr Opin Plant Biol. 11(1):23-7. Den Herder J, begeleider Temmerman W (2002) Studie van de fas virulentiegenen van Rhodococcus fascians. Licentiaatsthesis biotechnologie 2001-2002. Depuydt S, Doležal K, Van Lijsebettens M, Moritz T, Holsters M and Vereecke D (2008) Modulation of the hormone setting by Rhodococcus fascians results in ectopic KNOX activation in Arabidopsis thaliana. Plant Physiol. 146:1267-1281. Depuydt S, Putnam ML, Holsters M and Vereecke D (2006) Rhodococcus fascians, an emerging threat for ornamental crops. Floriculture, Ornamental and Plant Biotechnology: Advances and Topical Issues (1st Edition, Volume 5), Texeira da Silva JA (Ed), Global Science Books, Isleworth, UK, pp 480-489. Depuydt S (2008) Molecular biology of the Rhodococcus fascians-plant interaction. Ghent Univ., Belgium. Depuydt S, De Veylder L, Holsters M and Vereecke D (2008a) Eternal youth, the fate of developing Arabidopsis leaves upon infection with Rhodococcus fascians. Plant Physiol. Preview 10.1104/pp.108.131797. Depuydt S, Trenkamp A, Fernie S, Elfthieh JP, Renou M, Holsters M and Vereecle D (2008b) An integrated genomics approach to define niche establishment by Rhodococcus fascians. Plant Physiol. Preview 10.1104/pp.108.131805. Desomer J, Dhaese P and Van Montagu M (1988) Conjugative transfer of cadmium resistance plasmids in Rhodococcus fascians strains. J. Bacteriol. 70:2401-5. Desomer J, Dhaese P and Van Montagu M (1990) Transformation of Rhodococcus fascians by high-voltage electroporation and development of R. fascians cloning vectors. Appl. Environm. Microbiol. 56, 2818-2825. Dihanich ME, Najarian D, Clark R, Gillman EC, Martin NC and Hopper AK (1987) Isolation and characterization of MOD5, a gene required for isopentenylation of cytoplasmic and mitochondrial tRNAs of Saccharomyces cerevisiae. Mol Cell Biol. 7(1):177-84. Dobrev P, Motyka V, Gaudinova A, Malbeck JV, Travnickova A, Kaminek M and Vankova R (2002) Transient accumulation of cis- and trans-zeatin type cytokinins and its relation to cytokinin oxidase activity during cell cycle of synchronized tobacco BY-2 cells. Plant Physiol. Biochem. 40:333-337. Donderski W and Gluchowska M (2000) Production of Cytokinin-like Substances by Planktonic Bacteria Isolated from Lake Jeziorak. Polish Journal of Environmental Studies 9(5):369-376. Eason JR, Jameson PE and Bannister P (1995) Virulence assessment of Rhodococcus fascians strains on pea cultivars. Plant Pathol. 44:141-147. Eason JR, Morris RO and Jameson PE (1996) The relationship between virulence and cytokinin production by Rhodococcus fascians (Tilford 1936) Goodfellow 1984. Plant Pathol. 45:323-331. Einset JW and Skoog FK (1977) Isolation and identification of ribosyl-cis-zeatin from transfer RNA of Corynebacterium fascians. Biochem Biophys Res Commun. 79(4):1117-21. Emery RJ, Leport L, Barton JE, Turner NC and Atkins CA (1998) cis-isomers of cytokinins predominate in chickpea seeds throughout their development. Plant Physiol. 117(4):1515-23. Ericson JU and Björk GR (1986) Pleiotropic effects induced by modification deficiency next to the anticodon of tRNA from Salmonella typhimurium LT2. J Bacteriol. 166(3):1013-21. Esberg B, Leung HC E, Tsui HC T, Björk GR and Winkler ME (1999) Identification of the miaB gene, involved in methylthiolation of isopentenylated A37 derivatives in the tRNA of Salmonella typhimurium and Escherichia coli. J. Bacteriol. 181(23):7256-7265. Escobar MA and Dandekar AM (2003) Agrobacterium tumefaciens as an agent of disease.Trends Plant Sci. 8(8):380-6. References 159

Faivre-Amiot A (1967) Quelques observations sur la présence de Corynebacterium fascians (Tilford) Dowson dans les cultures maraichères et florales en France. Phytiatrie-Phytopharmacie 16:165-176. Feinbaum RL and Ausubel FM (1988) Transcriptional regulation of the Arabidopsis thaliana chalcone synthase gene. Mol Cell Biol 8:1985-1992. Francis I, Gevers D, Karimi M, Holsters M and Vereecke D (2007) Linear plasmids and phytopathogenicity. In Microbiology monographs 7: Microbial Linear Plasmids 99-115. Frébort I, Šebeka M, Galuszka P, Werner T, Schmülling T and Pe� P (2002) Cytokinin oxidase/cytokinin dehydrogenase assay: optimized procedures and applications. Anal Biochem 306:1-7. Frébortová J, Fraaije MW, Galuszka P, Sebela M, Pec P, Hrbác J, Novák O, Bilyeu KD, English JT and Frébort I (2004) Catalytic reaction of cytokinin dehydrogenase: preference for quinones as electron acceptors. Biochem J. 380:121-30. Galis I, Bilyeu K, Word G and Jameson PE (2005) Rhodococcus fascians: shoot proliferation without elevated cytokinins? Plant Growth Regulation 46:109-115. Gallegos MT, Schleif R, Bairoch A, Hofmann K and Ramos JL (1997) Arac/XylS family of transcriptional regulators. Microbiol Mol Biol Rev. 61(4):393-410. Galuszka P, Frébort I, Sebela M, Sauer P, Jacobsen S and Pec P (2001) Cytokinin oxidase or dehydrogenase? Mechanism of cytokinin degradation in cereals. Eur J Biochem. 268(2):450-61. Galuszka P, Frébortová J, Werner T, Yamada M, Strnad M, Schmülling T and Frébort I (2004) Cytokinin oxidase/dehydrogenase genes in barley and wheat: cloning and heterologous expression. Eur J Biochem. 271(20):3990-4002. Galuszka P, Popelková H, Werner T, Frébortová J, Mik V, Köllmer I, Schmülling T and Frébort I (2007) Biochemical characterization of cytokinin oxidases/dehydrogenases from Arabidopsis thaliana expressed in Nicotiana tabacum L. J Plant Growth Regul 26:255-267. Gaudin V, Vrain T and Jouanin L (1994) Bacterial genes modifying hormonal balances in plants. Plant Physiol. Biochem. 32(1):11-29. Georgiev MI, Pavlov AI and Bley T (2007) Hairy root type plant in vitro systems as sources of bioactive substances. Appl Microbiol Biotechnol 74:1175–1185. Gillissen B, Bürkle L, André B, Kühn C, Rentsch D, Brandl B and Frommer WB (2000) A new family of high-affinity transporters for adenine, cytosine, and purine derivatives in Arabidopsis. Plant Cell. 12(2):291-300. Giraud E, Moulin L, Vallenet D, Barbe V, Cytryn E, Avarre JC, Jaubert M, Simon D, Cartieaux F, Prin Y, Bena G, Hannibal L, Fardoux J, Kojadinovic M, Vuillet L, Lajus A, Cruveiller S, Rouy Z, Mangenot S, Segurens B, Dossat C, Franck WL, Chang WS, Saunders E, Bruce D, Richardson P, Normand P, Dreyfus B, Pignol D, Stacey G, Emerich D, Verméglio A, Médigue C and Sadowsky M (2007) Legumes symbioses: absence of Nod genes in photosynthetic bradyrhizobia. Science 316(5829):1307-12. Glover BJ, Torney K, Wilkins CG, Hanke DE (2008) CYTOKININ INDEPENDENT-1 regulates levels of different forms of cytokinin in Arabidopsis and mediates response to nutrient stress.� J Plant Physiol. 165(3):251-61. Goethals K, Van Montagu M and Holsters M (1992) Conserved motifs in a divergent nod box of Azorhizobium caulinodans ORS571 reveal a common structure in promoters regulated by LysR- Type proteins. PNAS 89:1646-1650. Goethals K, Vereecke D, Jaziri M, Van Montagu M and Holsters M (2001) Leafy gall formation by Rhodococcus fascians. Annu Rev Phytopathol. 39:27-52. References 160

Goldman BS, Nierman WC, Kaiser D, Slater SC, Durkin AS, Eisen JA, Ronning CM, Barbazuk WB, Blanchard M, Field C, Halling C, Hinkle G, Iartchuk O, Kim HS, Mackenzie C, Madupu R, Miller N, Shvartsbeyn A, Sullivan SA, Vaudin M, Wiegand R and Kaplan HB (2006) Evolution of sensory complexity recorded in a myxobacterial genome. Proc. Natl. Acad. Sci. U.S.A. 103(41):15200- 15205. Golovko A, Sitbon F, Tillberg E and Nicander B (2002) Identification of a tRNA isopentenyltransferase gene from Arabidopsis thaliana. Plant Mol Biol. 49(2):161-9. Goodfellow M (1984). Reclassification of Corynebacterium fascians (Tilford) Dowson in the genus Rhodococcus, as Rhodococcus fascians comb. nov. System. Appl. Microbiol. 5:225-229. Gray J, Gelvin SB, Meilan R and Morris RO (1996) Transfer RNA is the source of extracellular isopentenyladenine in a Ti-plasmidless strain of Agrobacterium tumefaciens. Plant Physiol. 110:431-438. Guo M, Manulis S, Barash I and Lichter A (2001) The operon for cytokinin biosynthesis of Erwinia herbicola pv. gypsophilae contains two promoters and is plant induced. Canadian Journal of Microbiology 47(12):1126-1131(6). Haberer G and Kieber JJ (2002) Cytokinins. New insights into a classic phytohormone. Plant Physiol. 128:354-362. Hakamada Y, Kobayashi T, Hitomi J, Kawai S and Ito S (1994) Molecular cloning and nucleotide sequence of the gene for an alkaline protease from the alkalophilic Bacillus sp. KSM-K16. J. Ferment. Bioeng. 78:105-108. Hake S, Smith HM, Holtan H, Magnani E, Mele G and Ramirez J (2004) The role of KNOX genes in plant development. Annu Rev Cell Dev Biol. 20:125-51. Hao G, Zhang H, Zheng D, Burr TJ (2005) luxR homolog avhR in Agrobacterium vitis affects the development of a grape-specific necrosis and a tobacco hypersensitive response. J Bacteriol. 187(1):185-92. Hao G and Burr TJ (2006) Regulation of long-chain N-acyl-homoserine lactones in Agrobacterium vitis. J Bacteriol. 188(6):2173-83. Hayashi Y, Onaka H, Itoh N, Seto H and Dairi T (2007) Cloning of the gene cluster responsible for biosynthesis of KS-505a (longestin), a unique tetraterpenoid. Biosci. Biotechnol. Biochem. 71(12):3072-3081. Helgeson JP and Leonard NJ (1966) Cytokinins: identification of compounds isolated from Corynebacterium fascians. Proc. Natl. Acad. Sci. USA 56:60-63. Heyl A and Schmülling T (2003) Cytokinin signal perception and transduction. Curr Opin Plant Biol. 6(5):480-8. Heyl A, Wulfetange K, Pils B, Nielsen N, Romanov GA, and Schmülling T (2007) Evolutionary proteomics identifies amino acids essential for ligand-binding of the cytokinin receptor CHASE domain. BMC Evol Biol. 17;7:62. Higuchi M, Pischke MS, Mähönen AP, Miyawaki K, Hashimoto Y, Seki M, Kobayashi M, Shinozaki K, Kato T, Tabata S, Helariutta Y, Sussman MR and Kakimoto T (2004) In planta functions of the Arabidopsis cytokinin receptor family. Proc Natl Acad Sci USA 101:8821-8826. Hirose N, Makita N, Yamaya T and Sakakibara H (2005) Functional characterization and expression analysis of a gene, OsENT2, encoding an equilibrative nucleoside transporter in rice suggest a function in cytokinin transport. Plant Physiol. 138(1):196-206. Hirose N, Takei K, Kuroha T, Kamada-Nobusada T, Hayashi H and Sakakibara H (2008) Regulation of cytokinin biosynthesis, compartmentalization and translocation. J Exp Bot. 59(1):75-83. Holub J, Hanuš J, Hanke DE and Strnad M (1998) Biological activity of cytokinins derived from ortho- and meta-hydroxybenzyladenine. Plant Growth Regul 26:109-115. Hou B, Lim EK, Higgins GS and Bowles DJ (2004) N-glucosylation of cytokinins by glycosyltransferases of Arabidopsis thaliana. J Biol Chem. 279(46):47822-32. References 161

Hunter WN (2007) The non-mevalonate pathway of isoprenoid precursor biosynthesis. J Biol Chem. 282(30):21573-7. Hutchison CE, Li J, Argueso C, Gonzalez M, Lee E, Lewis MW, Maxwell BB, Perdue TD, Schaller GE, Alonso JM, Ecker JR and Kieber JJ (2006) The Arabidopsis histidine phosphotransfer proteins are redundant positive regulators of cytokinin signaling. Plant Cell. 18(11):3073-87. Iacobellis NS, Sisto A, Surico G, Evidente A and DiMaio E (1994) Pathogenicity of Pseudomonas syringae subsp. savastanoi mutants defective in phytohormone production. J. Phytopathol. 140:238-248. Iacobellis NS, Caponero A and Evidente A (1998) Characterization of Pseudomonas syringae ssp. savastanoi strains isolated from ash. Plant Pathol. 47:73-83. Ibarra JA, Pérez-Rueda E, Segovia L and Puente JL (2008) The DNA-binding domain as a functional indicator: the case of the AraC/XylS family of transcription factors. Genetica 2008 133(1):65-76. Ihara M, Taya Y, Nishimura S, Tanaka Y (1984) Purification and some properties of delta 2- isopentenylpyrophosphate:5'AMP delta 2-isopentenyltransferase from the cellular slime mold Dictyostelium discoideum. Arch Biochem Biophys. 230(2):652-60. Ikeda H, Ishikawa J, Hanamoto A, Shinose M, Kikuchi H, Shiba T, Sakaki Y, Hattori M and Omura S (2003) Complete genome sequence and comparative analysis of the industrial microorganism Streptomyces avermitilis. Nat. Biotechnol. 21(5):526-531. Inoue T, Higuchi M, Hashimoto Y, Seki M, Kobayashi M, Kato T, Tabata S, Shinozaki K and Kakimoto T (2001) Identification of CRE1 as a cytokinin receptor from Arabidopsis. Nature 409:1060-1063. Ishii Y, Konishi J, Okada H, Hirasawa K, Onaka T and Suzuki M (2000) Operon structure and functional analysis of the genes encoding thermophilic desulfurizing enzymes of Paenibacillus sp. A11-2. Biochem. Biophys. Res. Commun. 270(1):81-88. Ishikawa J, Yamashita A, Mikami Y, Hoshino Y, Kurita H, Hotta K, Shiba T and Hattori M (2004) The complete genomic sequence of Nocardia farcinica IFM 10152. Proc. Natl. Acad. Sci. U.S.A. 101, 14925-14930. Isin EM and Guengerich FP (2007) Complex reactions catalyzed by cytochrome P450 enzymes. Biochim Biophys Acta. 1770(3):314-29. Jameson P (2000) Cytokinins and auxins in plant-pathogen interactions – An overview. Plant Growth Regulation 32(369-380). Jeon WB, Allard ST, Bingman CA, Bitto E, Han BW, Wesenberg GE and Phillips GN, Jr. (2006) X-ray crystal structures of the conserved hypothetical proteins from Arabidopsis thaliana gene loci At5g11950 and AT2g37210. Proteins 65:1051-1054. John MC and Amasino RM (1988) Expression of an Agrobacterium Ti plasmid gene involved in cytokinin biosynthesis is regulated by virulence loci and induced by plant phenolic compounds. J. Bacteriol. 170:790-795. Joshi MV and Loria R (2007) Streptomyces turgidiscabies possesses a functional cytokinin biosynthetic pathway and produces leafy galls. MPMI 20(7):751-758. Kakimoto T (2001) Identification of plant cytokinin biosynthetic enzymes as dimethylallyl diphosphate:ATP/ADP isopentenyltransferases. Plant Cell Physiol. 42(7):677-85. Kakimoto T (2003) Biosynthesis of cytokinins. J. Plant Res. 116:233-239. Karadeniz A, Topcuo�lu SF and Inan S (2006) Auxin, gibberellin, cytokinin and abscisic acid production in some bacteria. Worl Journal of Microbiol & Biotechn 22(10):1061-1064. Kasahara H, Takei K, Ueda N, Hishiyama S, Yamaya T, Kamiya Y, Yamaguchi S and Sakakibara H (2004) J Biol Chem. 279(14):14049-54. Kemp DR (1978) Indole-3-ylacetic acid metabolism of Corynebacterium fascians. Microbial Ecology, ed. MW Loutit, JAR Miles, p. 341-345. Berlin: Springer-Verlag. References 162

Kenelly MM, Cazorla FM, de Vicente A, Ramos C and Sundin GW (2007) Pseudomonas syringae Diseases of Fruit Trees: Progress Toward Understanding and Control. Plant Dis. 91:4-17. Kereszt A, Li D, Indrasumunar A, Nguyen CD, Nontachaiyapoom S, Kinkema M, Gresshoff PM (2007) Agrobacterium rhizogenes-mediated transformation of soybean to study root biology. Nat Protoc. 2007;2(4):948-52. Kers JA, Cameron KD, Joshi MV, Bukhalid RA, Morello JE, Wach MJ, Gibson DM and Loria R (2005) A large, mobile pathogenicity island confers plant pathogenicity on Streptomyces species. Mol. Microbiol. 55:1025-1033. Klämbt D, Thies G and Skoog F (1966) Isolation of cytokinins from Corynebacterium fascians. Proc. Natl. Acad. Sci. USA 56:52-59. Klämbt D, Holtz J, Helbach M and Maass H (1984) Die Biogenese der Cytokinine. Ber Dtsch Bot Ges 97:57-65. Klämbt D (1992) The biogenesis of cytokniins in higher plants: our present knowledge. In Physiology and Biochemistry of Cytokinins in Plants, ed. M Kaminek, DWS Mok, E Zazímalová, pp.25-27. The Hague: SPB Academic. Kloepper JWE and Schroth MN (1978) Plant growth-promoting rhizobacteria on radishes. In Proc. 4th internat. Conf. on plant pathogenic bacteria. A. Gibert-Clarey. (edit)INRA, Tours, pp 879-882. Kochetov GA and Sevostyanova (2005) Binding of the coenzyme and formation of the transketolase active center. IA IUBMB Life. 57(7):491-7. Koonin EV, Makarova KS and Aravind L (2001) Horizontal gene transfer in prokaryotes-qunatification and classification. Annu. Rev. Microbiol. 55:709-42. Krall L, Raschke M, Zenk MH, Baron C (2002) The Tzs protein from Agrobacterium tumefaciens C58 produces zeatin riboside 5'-phosphate from 4-hydroxy-3-methyl-2-(E)-butenyl diphosphate and AMP. FEBS Lett. 527(1-3):315-8. Kurakawa T, Ueda N, Maekawa M, Kobayashi K, Kojima M, Nagato Y, Sakakibara H, Kyozuka J (2007) Direct control of shoot meristem activity by a cytokinin-activating enzyme. Nature 445(7128):652-5. Kyozuka J (2007) Control of shoot and root meristem function by cytokinin. Curr Opin Plant Biol. 10(5):442-6. Lacey MS (1939) Studies on a bacterium associated with leafy galls, fasciations and "cauliflower" disease of various plants. Part III. Further isolations, inoculation experiments and cultural studies. Ann. Appl. Biol. 26:262-278. Lange BM, Wildung MR, McCaskill D and Croteau R (1998) A family of transketolases that directs isoprenoid biosynthesis via a mevalonate-independent pathway. Proc Natl Acad Sci U S A. 95(5):2100-4. Laureys F, Dewitte W, Witters E, Van Montagu M, Inzé D and Van Onckelen H (1998) Zeatin is indispensable for the G2-M transition in tobacco BY-2 cells. FEBS Lett. 426(1):29-32. Lavermicocca P, Lonigro SL, Valerio F, Evidente A and Visconti A (2002) Reduction of olive knot disease by a bacteriocin from Pseudomonas syringae pv. ciccaronei. Appl Environ Microbiol. 68(3):1403-7. LeChevalier HA (1986) Nocardioforms. In Bergey's Manual of Systematic Bacteriology, Vol. 2. Sneath PHA, Mair NS, Sharpe ME and Holt JG (eds). Baltimore: Williams & Wilkins, pp. 1458-1506. Lichter A, Barash I, Valinsky L and Manulis S (1995a) The genes involved in cytokinin biosynthesis in Erwinia herbicola pv. gypsophilae: characterization and role in gall formation. J. Bacteriol. 177:4457-4465. Lichter A, Manulis S, Sagee O, Gafni Y, Gray J, Meilan R, Morris RO and Barash I (1995b) Production of cytokinins by Erwinia herbicola pv. gypsophilae and isolation of a locus conferring cytokinin biosynthesis. Mol. Plant-Microbe Interact. 8:114-121. References 163

Lidstrom ME and Chistoserdova L (2002) Plants in the pink: cytokinin production by methylobacterium. J Bacteriol. 184(7):1818. López MA, Bannenberg G and Castresana C (2008) Controlling hormone signaling is a plant and pathogen challenge for growth and survival. Curr Opin Plant Biol. 11(4):420-7. Loria R, Kers J, Joshi M (2006) Evolution of plant pathogenicity in Streptomyces. Annu Rev Phytopathol. 44:469-87. Luehrsen K, Woese CR, Wolters J and Stackebrandt E (1989) Nucleotide sequence of the 5S rRNA from Rhodococcus fascians. Nucleic Acids Res. 17:7378. Maes T, Vereecke D, Ritsema T, Cornelis K, Thu HNT, Van Montagu M, Holsters M and Goethals K (2001) The att locus of Rhodococcus fascians strain D188 is essential for full virulence on tobacco through the production of an autoregulatory compound. MolecularMicrobiology 42(1):13-28. Mähönen AP, Higuchi M, Törmäkangas K, Miyawaki K, Pischke MS, Sussman MR, Helariutta Y and Kakimoto T (2006) Cytokinins regulate a bidirectional phosphorelay network in Arabidopsis. Curr Biol. 16(11):1116-22. Manulis S, Haviv-Chesner A, Brandl MT, Lindow SE and Barash I (1998) Differential involvement of indole-3-acetic acid biosynthetic pathways in pathogenicity and epiphytic fitness of Erwinia herbicola pv. gypsophilae. Mol. Plant-Microbe Interact. 11:634-642. Marco ML, Legac J and Lindow SE (2005) Pseudomonas syringae genes induced during colonization of leaf surfaces. Environ Microbiol 7:1379-1391. Martin JL and McMillan FM (2002) SAM (dependent) I AM: the S-adenosylmethionine-dependent methyltransferase fold. Curr Opin Struct Biol. 12(6):783-93. Martin RC, Mok MC and Mok DW (1999a) Isolation of a cytokinin gene, ZOG1, encoding zeatin O- glucosyltransferase from Phaseolus lunatus. Proc Natl Acad Sci U S A 96(1):284-9. Martin RC, Mok MC and Mok DW (1999b) A gene encoding the cytokinin enzyme zeatin O- xylosyltransferase of Phaseolus vulgaris. Plant Physiol. 120(2):553-8. Martin RC, Mok MC, Habben JE and Mok DW (2001) A maize cytokinin gene encoding an O- glucosyltransferase specific to cis-zeatin. Proc Natl Acad Sci U S A 98(10):5922-6. Maruyama A, Maeda M and Simidu U (1986) Occurrence of plant hormone (cytokinin)-producing bacteria in the sea. J Appl Microbiol 61(6):569-574. Mason MG, Li J, Mathews DE, Kieber JJ and Schaller GE (2004) Type-B response regulators display overlapping expression patterns in Arabidopsis. Plant Physiol. 135(2):927-37. Mason MG, Mathews DE, Argyros DA, Maxwell BB, Kieber JJ, Alonso JM, Ecker JR and Schaller GE (2005) Multiple type-B response regulators mediate cytokinin signal transduction in Arabidopsis. Plant Cell. 17(11):3007-18. Mathevon C, Pierrel F, Oddou JL, Garcia-Serres R, Blondin G, Latour JM, Ménage S, Gambarelli S, Fontecave M and Atta M (2007) tRNA-modifying MiaE protein from Salmonella typhimurium is a nonheme diiron monooxygenase. Proc Natl Acad Sci U S A 104(33):13295-300. Matsubara S, Armstrong DJ and Skoog F (1968) Cytokinins in tRNA of Corynebacterium fascians. Plant Physiol. 43(3):451-453. Matsubara S (1980) Structure-activity relationships of cytokinins. Phytochemistry 19:2239-2253. Matsubara T, Ohshiro T, Nishina Y and Izumi Y (2001) Purification, characterization, and overexpression of flavin reductase involved in dibenzothiophene desulfurization by Rhodococcus erythropolis D-1. Appl Environ Microbiol. 67(3):1179-84. McLean KJ, Sabri M, Marshall KR, Lawson RJ, Lewis DG, Clift D, Balding PR, Dunford AJ, Warman AJ, McVey JP, Quinn AM, Sutcliffe MJ, Scrutton NS and Munro AW (2005) Biodiversity of cytochrome P450 redox systems. Biochem Soc Trans. 33(Pt 4):796-801. Mergaert P, Van Montagu M and Holsters M (1997) Molecular mechanisms of Nod factor diversity. Mol Microbiol. 25(5):811-7. References 164

Miller JH (1972) Experiments in Molecular Genetics. New York: Cold Spring Harbor Laboratory. Miyawaki K, Matsumoto-Kitano M and Kakimoto T (2004) Expression of cytokinin biosynthetic isopentenyltransferase genes in Arabidopsis: tissue specificity and regulation by auxin, cytokinin, and nitrate. Plant J. 37(1):128-38. Miyawaki K, Tarkowski P, Matsumoto-Kitano M, Kato T, Sato S, Tarkowska D, Tabata S, Sandberg G and Kakimoto T (2006) Roles of Arabidopsis ATP/ADP isopentenyltransferases and tRNA isopentenyl-transferases in cytokinin biosynthesis. PNAS 103(44):16597-16603. Mok DW and Mok MC (2001) CYTOKININ METABOLISM AND ACTION. Annu Rev Plant Physiol Plant Mol Biol. 52:89-118. Mok MC, Martin RC, Dobrev PI, Vanková R, Ho PS, Yonekura-Sakakibara K, Sakakibara H and Mok DW (2005) Topolins and hydroxylated thidiazuron derivatives are substrates of cytokinin O- glucosyltransferase with position specificity related to receptor recognition. Plant Physiol. 137(3):1057-66 Mongodin EF, Shapir N, Daugherty SC, DeBoy RT, Emerson JB, Shvartzbeyn A, Radune D, Vamathevan J, Riggs F, Grinberg V, Khouri H, Wackett LP, Nelson KE and Sadowsky MJ (2006) Secrets of soil survival revealed by the genome sequence of Arthrobacter aurescens TC1. PLoS Genet. 2 (12), E214. Moriguchi K, Maeda Y, Satou M, Kataoka M, Tanaka N and Yoshida K (2000) Analysis of unique variable region of a plant root inducing plasmid, pRi1724, by the construction of its physical map and library.DNA Res.7(3):157-63. Müller B and Sheen J (2007) Advances in cytokinin signaling. Science 318(5847):68-9. Munro AW, Girvan HM and McLean KJ (2007) Cytochrome P450--redox partner fusion enzymes. Biochim Biophys Acta. 1770(3):345-59. Murai N, Skoog F, Doyle ME and Hanson RS (1980) Relationships between cytokinin production, presence of plasmids, and fasciation caused by strains of Corynebacterium fascians. Proc. Natl. Acad. Sci. USA 77:619-623. Murofushi N, Inoue A, Watanabe N, Yasua O, Takahashi N (1983) Identification of cytokinins in root exudate of the rice plant. Plant Cell Physiol 24: 87–92. Murray JD, Karas BJ, Sato S, Tabata S, Amyot L and Szczyglowski K (2007) A cytokinin perception mutant colonized by Rhizobium in the absence of nodule organogenesis. Science 315(5808):101- 4. Nishida H, Ohnishi Y, Beppu T and Horinouchi S (2007) Evolution of gamma-butyrolactone synthases and receptors in Streptomyces. Environ Microbiol. 9(8):1986-94. Nishimura C, Ohashi Y, Sato S, Kato T, Tabata S and Ueguchi C (2004) Histidine kinase homologs that act as cytokinin receptors possess overlapping functions in the regulation of shoot and root growth in Arabidopsis. Plant Cell. 16(6):1365-77. Nizan R, Barash I, Valinsky L, Lichter A and Manulis S (1997) The presence of hrp genes on the pathogenicity-associated plasmid of the tumorigenic bacterium Erwinia herbicola pv. gypsophilae. Mol. Plant-Microbe Interact. 10:677-682. Normand P, Lapierre P, Tisa LS, Gogarten JP, Alloisio N, Bagnarol E, Bassi CA, Berry AM, Bickhart DM, Choisne N, Couloux A, Cournoyer B, Cruveiller S, Daubin V, Demange N, Francino MP, Goltsman E, Huang Y, Kopp OR, Labarre L, Lapidus A, Lavire C, Marechal J, Martinez M, Mastronunzio JE, Mullin BC, Niemann J, Pujic P, Rawnsley T, Rouy Z, Schenowitz C, Sellstedt A, Tavares F, Tomkins JP, Vallenet D, Valverde C, Wall LG, Wang Y, Medigue C and Benson DR (2007) Genome characteristics of facultatively symbiotic Frankia sp. strains reflect host range and host plant biogeography. Genome Res. 17(1):7-15. Novak O, Tarkowski P, Tarkowska D, Dolezal K, Lenobel R and Strnad M (2003) Quantitative analysis of cytokinins in plants by liquid chromatography--single-quadrupole mass spectrometry. Anal. Chim. Acta 480:207-218. References 165

Oduro KA and Munnecke DE (1975) Persistence of pea cotyledons induced by Corynebacterium fascians. Phytopathology 65:1114-1116. Ohnishi Y, Ishikawa J, Hara H, Suzuki H, Ikenoya M, Ikeda H, Yamashita A, Hattori M and Horinouchi S (2008) Genome sequence of the streptomycin-producing microorganism Streptomyces griseus IFO 13350. J. Bacteriol. 190(11):4050-4060. Oliynyk M, Samborskyy M, Lester JB, Mironenko T, Scott N, Dickens S, Haydock SF and Leadlay PF (2007) Complete genome sequence of the erythromycin-producing bacterium Saccharopolyspora erythraea NRRL23338. Nat. Biotechnol. 25(4):447-453. de O Manes CL, Van Montagu M, Prinsen E, Goethals K and Holsters M (2001) De novo cortical cell division triggered by the phytopathogen Rhodococcus fascians in tobacco. Mol Plant Microbe Interact. 14(2):189-95. de O Manes CL, Beeckman T, Ritsema T, Van Montagu M, Goethals K and Holsters M (2004) Phenotypic alterations in Arabidopsis thaliana plants caused by Rhodococcus fascians infection. J Plant Res 117:139-145. Pallen MJ and Wren BW (2007) Bacterial pathogenomics. Nature. 449(7164):835-42. Peng J and Harberd NP (1997) Gibberellin deficiency and response mutations suppress the stem elongation phenotype of phytochrome-deficient mutants of Arabidopsis. Plant Physiol 113:1051-1058. Peng J, Peng F, Zhu C and Wei S (2008) Molecular cloning of a putative gene encoding isopentenyltransferase from pingyitiancha (Malus hupehensis) and characterization of its response to nitrate. Tree Physiol. 28(6):899-904. Penyalver R, García A, Ferrer A, Bertolini E, Quesada JM, Salcedo CI, Piquer J, Pérez-Panadés J, Carbonell EA, Del Río C, Caballero JM and López MM (2006) Factors Affecting Pseudomonas savastanoi pv. savastanoi Plant Inoculations and Their Use for Evaluation of Olive Cultivar Susceptibility. Phytopathology 96(3):313-9. Pérez-Martínez I, Zhao Y, Murillo J, Sundin GW and Ramos C (2008) Global genomic analysis of Pseudomonas savastanoi pv. savastanoi plasmids. J Bacteriol. 190(2):625-35. Persson BC and Björk GR (1993) Isolation of the gene (miaE) encoding the hydroxylase involved in the synthesis of 2-methylthio-cis-ribozeatin in tRNA of Salmonella typhimurium and characterization of mutants. J Bacteriol 175(24):7776-85. Pierrel F, Hernandez HL, Johnson MK, Fontecave M and Atta M (2003) MiaB protein from Thermotoga maritima. Characterization of an extremely thermophilic tRNA-methylthiotransferase. J Biol Chem. 278(32):29515-24. Pierrel F, Douki T, Fontecave M and Atta M (2004) MiaB protein is a bifunctional radical-S- adenosylmethionine enzyme involved in thiolation and methylation of tRNA. J Biol Chem. 279(46):47555-63. Pisabarro A, Correia A and Martin JF (1998) Pulse-field gel electrophoresis analysis of the genome of Rhodococcus fascians: genome size and linear and circular replicon compostion in virulent and avirulent strains. Curr. Microbiol. 36:302-8. Phan AP, Ngo TT and Lenhoff HM (1982) Spectrophotometric assay for lysine decarboxylase. Anal Biochem. 120(1):193-7. Phillips DA and Torrey JG (1972) Studies on Cytokinin Production by Rhizobium. Plant Physiol. 49(1):11-15. Popelková H, Fraaije MW, Novák O, Frébortová J, Bilyeu KD and Frébort I (2006) Kinetic and chemical analyses of the cytokinin dehydrogenase-catalysed reaction: correlations with the crystal structure. Biochem J. 398(1):113-24. Powell GK and Morris RO (1986) Nucleotide sequence and expression of a Pseudomonas savastanoi cytokinin biosynthetic gene: homology with Agrobacterium tumefaciens tmr and tzs loci. Nucl. Acids Res. 14:2555-2565. References 166

Prinsen E, Kamínek M and Van Onckelen HA (1997) Cytokinin biosynthesis: a black box? Plant Growth Regul 23:3-15. Putnam ML and Miller ML (2007) Rhodococcus fascians in herbaceous perennials. Plant disease 91(9):1064-1076. Rahtbone MP and Hall RH (1972) Concerning the presence of the cytokinin N6-(∆2- isopentenyl)adenine in cultures of Corynebacterium fascians. Planta 108:93-102. Rajkowitsch L, Chen D, Stampfl S, Semrad K, Waldsich C, Mayer O, Jantsch MF, Konrat R, Bläsi U, and Schroeder R (2007) RNA chaperones, RNA annealers and RNA helicases. RNA Biol. 4(3):118-30. Rashotte AM, Mason MG, Hutchison CE, Ferreira FJ, Schaller GE and Kieber JJ (2006) A subset of Arabidopsis AP2 transcription factors mediates cytokinin responses in concert with a two-component pathway. Proc Natl Acad Sci U S A. 103(29):11081-5. Raven JA and Allen JF (2003) Genomics and chloroplast evolution: what did cyanobacteria do for plants? Genome Biol. 4(3):209. Regier DA, Akiyoshi DE and Gordon MP (1989) Nucleotide sequence of the tzs gene from Agrobacterium rhizogenes strain A4. Nucleic Acids Res. 17(21):8885. Riefler M, Novak O, Strnad M and Schmülling T (2006) Arabidopsis cytokinin receptor mutants reveal functions in shoot growth, leaf senescence, seed size, germination, root development, and cytokinin metabolism. Plant Cell 18:40-54. Rinaldi AC and Comandini O (1999) Cytokinin oxidase: new insights into enzyme properties. Trends Plant Science 4(4):127-128. Robert-Seilaniantz A, Navarro L, Bari R and Jones JD (2007) Pathological hormone imbalances. Curr Opin Plant Biol. 10(4):372-379. Rocak S and Linder P (2004) DEAD-box proteins: the driving forces behind RNA metabolism. Nature reviews Molecular cell biology 5:232-241. Rohmer M (2003) Mevalonate-independent methylerythritol phosphate pathway for isoprenoid biosynthesis. Elucidation and distribution. Pure Appl. Chem. 75(Nos. 2-3) :375-387. Romanov GA, Lomin SN and Schmülling T (2006) Biochemical characteristics and ligand-binding properties of Arabidopsis cytokinin receptor AHK3 compared to CRE1/AHK4 as revealed by a direct binding assay. J Exp Bot. 57(15):4051-8. Roussaux J (1965) Etude préliminaire des modifications induites chez le pois express Alaska par le Corynebacterium fascians (Tilford) Dowson. Rev. Gén. Bot. 72:21-53. Roussaux J (1975) Stimulation and inhibition reactions in plants infected by Corynebacterium fascians (Tilford) Dowson. Marcellia 38:305-310. Sakakibara H, Kasahara H, Ueda N, Kojima M, Takei K, Hishiyama S, Asami T, Okada K, Kamiya Y, Yamaya T and Yamaguchi S (2005) Agrobacterium tumefaciens increases cytokinin production in plastids by modifying the biosynthetic pathway in the host plant. Proc Natl Acad Sci U S A 102(28):9972-7. Sakakibara H (2006) Cytokinins: activity, biosynthesis, and translocation. Annu. Rev. Plant Biol. 57:431-449. Sambrook J, Fritsch EF and Maniatis T (1989) Molecular cloning: a laboratory manual. Cold spring harbor laboratory press, cold spring harbor, NY. Sakamoto T, Sakakibara H, Kojima M, Yamamoto Y, Nagasaki H, Inukai Y, Sato Y and Matsuoka M (2006) Ectopic expression of KNOTTED1-like homeobox protein induces expression of cytokinin biosynthesis genes in rice. Plant Physiol. 142(1):54-62. Sakano Y, Okada Y, Matsunaga A, Suwama T, Kaneko T, Ito K, Noguchi H and Abe I (2004) Molecular cloning, expression, and characterization of adenylate isopentenyltransferase from hop (Humulus lupulus L.). Phytochemistry 65:2439-2446. References 167

Scarbrough E, Armstrong DJ, Skoog F, Frihart CR and Leonard NJ (1973) Isolation of cis-zeatin from Corynebacterium fascians cultures. Proc. Natl. Acad. Sci. USA 70:3825-3829. Schmülling T, Werner T, Riefler M, Krupková E and Bartrina y Manns I (2003) Structure and function of cytokinin oxidase/dehydrogenase genes of maize, rice, Arabidopsis and other species. J Plant Res. 116(3):241-52. Schubert HL, Blumenthal RM and Cheng X (2003) Many paths to methyltransfer: a chronicle of convergence. Trends Biochem Sci. 28(6):329-35. Shigaki T, Nelson SC and Alvarez AM (2000) Symptomless Spread of Blight-inducing Strains of Xanthomonas campestris pv. campestris on Cabbage Seedlings in Misted Seedbeds. European Journal of Plant Pathology, 106(4):339-346(8). Shirai T, Suzuki A, Yamane T, Ashida T, Kobayashi T, Hitomi J and Ito S (1997)High-resolution crystal structure of M-protease: phylogeny aided analysis of the high-alkaline adaptation mechanism. Protein Eng. 106):627-634. Shoeb F, Yadav JS, Bajaj S and Rajam MV (2001) Polyamines as biomarkers for plant regeneration capacity: improvement of regeneration by modulation of polyamine metabolism in different genotypes of indica rice. Plant Science 160:1229-1235. Sisto A, Cipriano MG and Morea M (2004) Knot Formation Caused by Pseudomonas syringae subsp. savastanoi on Olive Plants Is hrp-Dependent. Phytopathology 94(5):484-489. Spaepen S, Vanderleyden J and Remans R (2007) Indole-3-acetic acid in microbial and micro- organism-plant signaling. FEMS Microbiol Rev 31:425-448. Spatz K, Köhn H and Redenbach M (2002) Characterization of the Streptomyces violaceoruber SANK95570 plasmids pSV1 and pSV2. FEMS Microbiology Letters 213:87-92. Spena A, Estruch JJ and Schell J (1992) On microbes and plants: new insights in phytohormonal research. Curr. Opin. Biotech. 3:159-163. Spíchal L, Rakova NY, Riefler M, Mizuno T, Romanov GA, Strnad M and Schmülling T (2004) Two cytokinin receptors of Arabidopsis thaliana, CRE1/AHK4 and AHK3, differ in their ligand specificity in a bacterial assay. Plant Cell Physiol 45:1299-1035. Srivastava S and Srivastava AK (2007) Hairy root culture for mass-production of high-value secondary metabolites. Crit Rev Biotech 27:29-43. Stange RR Jr, Jeffares D, Young C, Scott DB, Eason JR and Jameson PE. (1996) PCR amplification of the fas-1 gene for the detection of virulent strains of Rhodococcus fascians. Plant Pathol. 45:407-417. Stevens GA Jr. and Berry AM (1988) Cytokinin Secretion by Frankia sp. HFP ArI3 in Defined Medium. Plant Physiology 87:15-16. Stinear TP, Seemann T, Pidot S, Frigui W, Reysset G, Garnier T, Meurice G, Simon D, Bouchier C, Ma L, Tichit M, Porter JL, Ryan J, Johnson PD, Davies JK, Jenkin GA, Small PL, Jones LM, Tekaia F, Laval F, Daffe M, Parkhill J and Cole ST (2007) Reductive evolution and niche adaptation inferred from the genome of Mycobacterium ulcerans, the causative agent of Buruli ulcer. Genome Res. 17(2):192-200. Stirk WA, Novák O, Václavíková K, Tarkowski P, Strnad M and van Staden J Planta (2008) Spatial and temporal changes in endogenous cytokinins in developing pea roots. 227(6):1279-89. Strnad M, Vanék T, Binarová P, Kaminek M and Hanuš J (1990) Enzyme immunoassays for cytokinins in alfalfa cell culture in: Kutá�ek, M., Elliott, M. C., Machá�ková, J. (Eds.), Molecular aspects of hormonal regulation of plant development, SPB Academic Publ., The Hague, 1990, pp. 41 – 54. Sturtevant DB and Taller B (1989) Cytokinin production by Bradyrhizobium japonicum. Plant Physiol. 89:1247-1252. Sun J, Niu QW, Tarkowski P, Zheng B, Tarkowska D, Sandberg G, Chua NH and Zuo J (2003) The Arabidopsis AtIPT8/PGA22 gene encodes an isopentenyl transferase that is involved in de novo cytokinin biosynthesis. Plant Physiol. 131(1):167-76. References 168

Surico G, Iacobellis NS and Sisto A (1985) Studies of the role of indole-3-acetic acid and cytokinins in the formation of knots on olive and oleander plants by Pseudomonas syringae pv. savastanoi. Physiol. Plant Pathol. 20:309-320. Suzuki T, Sakurai K, Imamura A, Nakamura A, Ueguchi C and Mizuno T (2000) Compilation and characterization of histidine-containing phosphotransmitters implicated in His-to-Asp phosphorelay in plants: AHP signal transducers of Arabidopsis thaliana. Biosci Biotechnol Biochem. 64(11):2486-9 Suzuki T, Miwa K, Ishikawa K, Yamada H, Aiba H and Mizuno T (2001) The Arabidopsis sensor His-kinase, AHK4, can respond to cytokinins. Plant Cell Physiol 42:107-113. Takei K, Sakakibara H and Sugiyama T (2001) Identification of genes encoding adenylate isopentenyltransferase, a cytokinin biosynthesis enzyme, in Arabidopsis thaliana. J. Biol. Chem. 276: 26405-26410. Takei K, Dekishima Y, Eguchi T, Yamaya T, Sakakibara H (2003)� A new method for enzymatic preparation of isopentenyladenine-type and trans-zeatin-type cytokinins with radioisotope-labeling. J Plant Res. 116(3):259-63. Takei K, Yamaya T and Sakakibara H (2004)� Arabidopsis CYP735A1 and CYP735A2 encode cytokinin hydroxylases that catalyze the biosynthesis of trans-Zeatin. J Biol Chem. 279(40):41866- 72. Taller BJ and Wong TY (1989) Cytokinins in Azotobacter vinelandii Culture Medium. Appl Environ Microbiol. 55(1):266-267. Tanaka M, Takei K, Kojima M, Sakakibara H and Mori H (2006)� Auxin controls local cytokinin biosynthesis in the nodal stem in apical dominance. Plant J. 45(6):1028-36. Tang X, Xiao Y and Zhou JM (2006)�Regulation of the type III secretion system in phytopathogenic bacteria. Mol Plant Microbe Interact. 19(11):1159-66. Tao M, Wang L, Wendt-Pienkowski E, George NP, Galm U, Zhang G, Coughlin JM and Shen B (2007) The tallysomycin biosynthetic gene cluster from Streptoalloteichus hindustanus E465-94 ATCC 31158 unveiling new insights into the biosynthesis of the bleomycin family of antitumor antibiotics. Mol Biosyst 3(1):60-74. Tarkowská D, Dole al K, Tarkowski P, Åstot C, Holub J, Fuksová K, Schmülling T, Sandberg G and Strnad M (2003) Identification of new aromatic cytokinins in Arabidopsis thaliana and Populus × canadensis leaves by LC-(+)ESI-MS and capillary liquid chromatography/frit–fast atom bombardment mass spectrometry. Physiologia plantarum 117(4):579-590. Taya Y, Tanaka Y and Nishimura S (1978) 5'-AMP is a direct precursor of cytokinin in Dictyostelium discoideum. Nature 271: 545-547. Temmerman W, Vereecke D, Dreesen R, Van Montagu M, Holsters M and Goethals K (2000) Leafy gall formation is controlled by fasR, an AraC-type regulatory gene in Rhodococcus fascians. J Bacteriol. 182(20):5832-40. Temmerman W (2000) Role of the fas locus in leafy gall development by the phytopathogen Rhodococcus fascians. Ghent Univ., Belgium. Thimann KV and Sachs T (1966) The role of cytokinins in the “fasciation” disease caused by Corynebacterium fascians. Amer. J. Bot. 53(7):731-739. Tien TM, Gaskins MH and Hubbell DH (1979) Plant Growth Substances Produced by Azospirillum brasilense and Their Effect on the Growth of Pearl Millet (Pennisetum americanum L.). Appl Environ Microbiol. 37(5):1016-1024. Tilford PE (1936) Fasciation of sweet peas caused by Phytomonas fascians n. sp. J. Agric. Res. 53:383-394. To JPC and Kieber JJ (2007) Cytokinin signaling: two-components and more. Trends Plant Sci. 13:85- 92. References 169

To JP, Deruère J, Maxwell BB, Morris VF, Hutchison CE, Ferreira FJ, Schaller GE and Kieber JJ (2007) Cytokinin regulates type-A Arabidopsis Response Regulator activity and protein stability via two-component phosphorelay. Plant Cell. 19(12):3901-14. Tsavkelova EA, Klimova SIu, Cherdyntseva TA and Netrusov AI (2006) Microbial producers of plant growth stimulators and their practical use: A review. Applied Biochemistry and Microbiology 42 (2): 117- 126 Russian Prikl Biokhim Mikrobiol. 42(2):133-143. Ueguchi C, Sato S, Kato T and Tabata S (2001) The AHK4 gene involved in the cytokinin-signaling pathway as a direct receptor molecule in Arabidopsis thaliana. Plant Cell Physiol. 42(7):751-5. Vandeputte O, Öden S, Mol A, Vereecke D, Goethals K, El Jaziri M and Prinsen E (2005) Biosynthesis of auxin by the gram-positive phytopathogen Rhodococcus fascians is controlled by compounds specific to infected plant tissues. Appl. Environ. Microbial. 71(3):1169-1177. Vandeputte O, Vereecke D, Mol A, Lenjou M, Van Bockstaele D, El Jaziri M and Baucher M (2007) Rhodococcus fascians infection accelerates progression of tobacco BY-2 cells into mitosis through rapid changes in plant gene expression. New Phytol. 175(1):140-54. Veach YK, Martin RC, Mok DWS, Malbeck J, Vankova R and Mok MC (2003) O-glucosylation of cis- zeatin in maize. Characterization of genes, enzymes, and endogenous cytokinins. Plant Physiol. 131:1374-1380. Vereecke D, Burssens S, Simón-Mateo C, Inzé D, Van Montagu M, Goethals K and Jaziri M (2000) The Rhodococcus fascians-plant interaction: morphological traits and biotechnological applications. Planta 210:241-251. Vereecke D, Cornelis K, Temmerman W, Jaziri M, Van Montagu M, Holsters M and Goethals K (2002) Chromosomal locus that affects pathogenicity of Rhodococcus fascians. Journal of Bacteriology 184:1112-1120. Vereecke D, Temmerman W, Jaziri M, Holsters M and Goethals K (2003) Toward an understanding of the Rhodococcus fascians-plant interaction. In Plant-Microbe Interactions 6:53-79. Edited by Stacey G and Keen NT, APS press. Vitorello CBM, Camargo LEA, Sluys MAV, Kitajima JP, Truffi D, Amaral AM, Harakava R, Franco JC, Wood D, Oliveira MC, Miyaki C, Takita MA, Silva ACR, Furlan LR, Carraro DM, Camarotte G, Almeida NF, Carrer H, Coutinho LL, El-Dorry HA, Ferro MIT, Gagliardi PR, Giglioti E, Goldman MHS, Goldman GH, Kimura ET, Ferro ES, Kuramae EE, Lemos EGM, Lemos MVF, Mauro SMZ, Machado MA, Marino CL, Menck CF, Nunes LR, Oliveira RC, Pereira GG, Siqueira W, Souza AA, Tsai SM, Zanca AS, Simpson AJG, Brumbley SM and Setubal JC (2004) The genome sequence of the gram-positive sugarcane pathogen Leifsonia xyli subsp. xyli. Mol. Plant Microbe Interact. 17(8):827-836. Wang Y, Luo JP, Wei ZJ and Zhang JC (2008) Molecular cloning and expression analysis of a cytokinin oxidase (DhCKX) gene in Dendrobium huoshanense. Mol Biol Rep. DOI 10.1007/s11033-008-9316- 2. Werck-Reichhart D and Feyereisen R (2000) Cytochromes P450: a success story. Genome Biol. 1(6):REVIEWS3003.1-3003.9. Werner T, Motyka V, Strnad M and Schmülling T (2001) Regulation of plant growth by cytokinin. Proc Natl Acad Sci U S A 98(18):10487-92. Werner T , Motyka V, Lacou V, Smets R, Van Onckelen H and Schmülling T (2003) Cytokinin-deficient transgenic Arabidopsis plants show multiple developmental alterations indicating opposite functions of cytokinins in the regulation of shoot and root meristem activity. Plant Cell 15:2532-2550. Werner T, Köllmer I, Bartrina I, Holst K and Schmülling T (2006) New insights into the biology of cytokinin degradation. Plant Biol (Stuttg) 8(3):371-81. Weinthal DM, Barash I, Panijel M, Valinsky L, Gaba V and Manulis-Sasson S (2007) Distribution and replication of the pathogenicity plasmid pPATH in diverse populations of the gall-forming bacterium Pantoea agglomerans. Appl Environ Microbiol. 73(23):7552-61. References 170

Yamada H, Suzuki T, Terada K, Takei K, Ishikawa K, Miwa K, Yamashino T and Mizuno T (2001) The Arabidopsis AHK4 histidine kinase is a cytokinin-binding receptor that transduces cytokinin signals across the membrane. Plant Cell Physiol 42:1017-1023. Yang SH, Yu H and Goh CJ (2003) Functional characterisation of a cytokinin oxidase gene DSCKX1 in Dendrobium orchid. Plant Mol Biol. 51(2):237-48. Ye C, Wu S, Kong F, Zhou C, Yang Q, Sun Y and Wang B (2006) Identification and characterization of an isopentenyltransferase (IPT) gene in soybean (Glycine max L.). Plant Science 170:542-550. Yonekura-Sakakibara K, Kojima M, Yamaya T, Sakakibara H (2004) Molecular characterization of cytokinin-responsive histidine kinases in maize. Differential ligand preferences and response to cis- zeatin. Plant Physiol 134:1654-1661. Young JP, Crossman LC, Johnston AW, Thomson NR, Ghazoui ZF, Hull KH, Wexler M, Curson AR, Todd JD, Poole PS, Mauchline TH, East AK, Quail MA, Churcher C, Arrowsmith C, Cherevach I, Chillingworth T, Clarke K, Cronin A, Davis P, Fraser A, Hance Z, Hauser H, Jagels K, Moule S, Mungall K, Norbertczak H, Rabbinowitsch E, Sanders M, Simmonds M, Whitehead S and Parkhill (2006) The genome of Rhizobium leguminosarum has recognizable core and accessory components. J.Genome Biol. 7(4):R34. Zhao Y (2008) The role of local biosynthesis of auxin and cytokinin in plant development. Curr Opin Plant Biol. 11(1):16-22. Zhu J, Oger PM, Schrammeijer B, Hooykaas PJJ, Farrand SK and Winans SC (2000) The bases of crown gall tumorigenesis. J Bacteriol. 182(14):3885-95. Zubko E, Adams CJ, Macháèková I, Malbeck J, Scollan C, Meyer P (2002) Activation tagging identifies a gene from Petunia hybrida responsible for the production of active cytokinins in plants. Plant J. 29(6):797-808. Zupan JR and Zambryski P (1995) Transfer of T-DNA from Agrobacterium to the plant cell. Plant Physiol. 107:1041-1047.

Dankwoord

Aan alles komt een eind en zo ook aan deze doctoraatsstudie. Vier jaar: het lijkt lang, maar toch is het net gisteren dat ik mijn project voor de jury van het IWT ging verdedigen. Vier jaar waarin ik zoveel nieuwe ervaringen heb opgedaan en de microbe voor wetenschappelijk onderzoek alleen maar aangewakkerd is. Vier jaar die gepaard gingen met gejuich voor elk gelukt experiment, maar ook soms luid gevloek wanneer het niet wou verlopen zoals het hoorde. Een periode waarin ik veel nieuwe mensen heb leren kennen en waarin de vertrouwde mensen mij door dik en dun gesteund hebben. De laatste maanden zijn er toch ook een beetje kuren, zweet en tranen gepaard gegaan met het schrijven van deze thesis, maar eindelijk kan ik beginnen aan het schrijven van het laatste stuk: mijn dank aan iedereen die me tot dit punt gebracht heeft. De mensen die me goed kennen weten dat het er niet zo vlot uit rolt, maar ik ga mijn best doen. Koen, ik wil je bedanken om in mij te geloven en erop te vertrouwen dat ik het fas-verhaal tot een goed einde kon brengen. Ik ben ervan overtuigd dat de discussies en het afvuren van de vele vragen mij hebben geholpen om mijn IWT-beurs binnen te halen. Marcelle, bedankt om de taak als promotor over te nemen. Merci voor de steun, suggesties en je enthousiasme. Danny, ik denk niet dat ik je genoeg kan bedanken voor de voorbije vier jaar. Bedankt voor alle adviezen, begeleiding, discussies, het lezen van mijn schrijfsels, enthousiasme, positivisme en nog veel meer. Bedankt dat je mij onder je vleugels genomen hebt. Ook heel erg merci aan mijn collega’s Rhodococcen. Isolde, het zit erop, we zijn er geraakt. Bedankt voor de voorbije vier jaar, zowel in het labo als daarbuiten. Stephen en Elisabeth: bedankt voor de sfeer in onze mini-groep. Elisabeth, jij bent de laatste Rhodococ, maar zeker niet de minste. Veel succes nog met al je werk, het komt zeker goed! Ook de overige ‘Plant- Microbes’ en ‘ex-Plant-Microbes’ wil ik zeker bedanken voor alle hulp en de goeie sfeer: Annick, Christa, Sofie, Virginie, Katrien, Ward, Willem, Griet, Juan Carlos en Jeroen. Chris en Geert, bedankt om mij zoveel uit de nood te helpen met de HPLC. Raf, heel erg bedankt om mij te helpen met de computertechnologische kantjes van deze thesis. En natuurlijk ook bedankt aan de vele andere mensen in het departement voor alle hulp en de sfeer. Tine, ik ga onze koffiepauzes missen. De eindmeet is in zicht. Nog veel courage en succes! Laurens, je bent er ook bijna. Veel succes nog. En we gaan zeker nog een keer een pintje drinken! I also want to thank Prof. Miroslav Strnad for giving me the opportunity to work in the lab in Olomouc. Petr (Tarkowski), Lukas, and Petr (Galuszka): thanks for all of your help to unravel the Rhodococcus cytokinin mystery. And last but not least, a very big thank you also to Katka! When I think about our sometimes endless discussions I get a smile on my face. Thanks for your patience for teaching me the techniques and for the good time in Olomouc! En natuurlijk zijn er nog mijn vrienden en familie, die me door dik en dun gesteund hebben. Heel erg bedankt aan allemaal. Oneindig veel dank ook voor mijn ouders. Pa en moetje, bedankt dat jullie altijd klaarstaan voor mij; voor jullie steun en raad; gewoon bedankt voor alles. Stef, keppie, jij bent de laatste in de rij. De laatste maanden waren vrij intens, maar we hebben het overleefd. Heel erg merci voor het doorstaan van al mijn kuren, voor al je hulp en je onvoorwaardelijke steun. Gewoon bedankt om er te zijn. Ik zie je graag!

Ine.

CURRICULUM VITAE

Personalia Name: Pertry Given names: Ine Home Address: Komenseweg 31 8902 Zillebeke (Ieper) Tel.: 0498/105923 Date and Place of birth: 11 juni 1982, Ieper Age: 26 e-mail: [email protected] Nationalityt: Belgian Marital state: single

Education

1994-2000 Undergraduate Sciences-Math, Koninklijk Atheneum Ieper 2000-2004 Graduate (Licentiate) Biotechnology, Universityt Ghent 2004-2008 PhD in Sciences: Biotechnologie, University Ghent

List of Publications

In a Refereed Journal Pertry I, Václavíková K, Depuydt S, Galuszka P, Spíchal L, Temmerman W, Stes E, Schmülling T, Kakimoto T, Van Montagu MCE, Strnad M, Holsters M, Tarkowski P, and Vereecke D. Identification of Rhodococcus fascians cytokinins and their modus operandi to reshape the plant. Proc Natl Acad Sci U S A. 2009 Jan 20;106(3):929-34.

In preparation Pertry I, Václavíková K, Galuszka P, Spíchal L Depuydt S, Temmerman W, Riefler M, Schmülling T, Strnad M, Holsters M, Tarkowski P, and Vereecke D. Biochemistry and biology of Rhodococcus fascians cytokinin biosynthesis. Pertry I, De Backer P, Holsters M and Vereecke D. Bacterial cytokinin production: different pathways for different purposes.

Lectures

Pertry I. How Rhodococcus fascians reshapes the plant: towards the identification of the fas molecules. Palacky University, Olomouc, Czech Republic (14.06.2007).

Participation at courses and meetings

With communication or poster: Pertry I, Holsters M and Vereecke D. How Rhodococcus fascians reshapes the plant: towards the identification of the fas molecules. Abstract presented at the 13th International Congress on Molecular Plant-Microbe Interactions (21.07.2007-27.07.2007, Sorrento, Italy). Stes E, Depuydt S, Francis I, Pertry I, Holsters M and Vereecke D. Rhodococcus fascians: an emerging threat in the ornamentals industry. Abstract presented at the VIB seminar (01.03.2007-02.03.2007, Blankenberge). Vereecke D, Francis I, Pertry I, Depuydt S, Stes E and Holsters M. The interaction between the phytopathogenic Actinomycete Rhodococcus fascians and its host plants. Abstract presented at the Journées Actinomycètes 2006, Ecully (France), 14-15 June 2006. Francis I, Pertry I, Depuydt S, Vereecke D, Holsters M and Goethals K. The phytopathogenic Actinomycete Rhodococcus fascians as a model system to study adaptation to endophytic growth and induced alterations of host development. Abstract presented at the VIB seminar (3.03.2005 – 4.03.2005, Blankenberge).

Participant: � VIB seminar (11.03.2004-12.03.2004, Blankenberge) � Root symposium (28.11.2005, Gent) � VIB Science Club Host-Pathogen Interactions (9.01.06, Leuven) � VIB seminar (9.03.2006 – 10.03.2006, Blankenberge) � Mini-symposium Biofuels (28.11.06, Gent)

Workshops: � Plant Cell Cycle workshop (15.03.2005, Gent) � Effective writing and presenting for life-sciences research (3-4.10.2007, Gent)

Fellowships

01.10.2004-30.09.2006 IWT bursar 01.10.2006-30.09.2008 IWT bursar 12.06.2007-25.06.2007 EMBO Short Term Fellowship

Experience

2004-2008 IWT-bursar, (Plant Biotechnology and Genetics/Ghent University; Department of Plant Systems Biology/VIB2

Research of the plant-Rhodococcus fascians interactions by standard genetical, microbial, molecular, biochemical and protein techniques

Promotor: Prof. Dr. Marcelle Holsters, co-promotor: Dr. Danny Vereecke.

Study periods abroad 12.06.2007-25.06.2007, 04.11.2007-14.12.2007 Department Biochemistry, Palacky University, Slechtitelu 11, Olomouc, CZ-783 71 Czech Republic Purification of cytokinins via immuno chromatography and HPLC.

Teaching experience

Theoretical courses (Ghent University)

2006-2007 Guest lecturer of “ Moleculaire grondslagen in de biotechnologie”, Prof. Holsters

2007-2008 Guest lecturer of “ Moleculaire grondslagen in de biotechnologie”, Prof. Holsters

Practical courses (Ghent University) 2005-2006 Guidance of undergraduates in Biology (2nd bachelor) Practicum Genetics and Molecular Techniques 2006-2007 Guidance of undergraduates in Biology (2nd bachelor) Practicum Genetics and Molecular Techniques 2007-2008 Guidance of undergraduates in Biology (2nd bachelor) Practicum Genetics and Molecular Techniques

Guidance of students 2005-2006 "Molecular analysis of the regulation of fas gene expression”, Carmen de Sena Tomás (Socrates student) 2006-2007 “De interactie tussen Rhodococcus fascians en zijn gastheerplanten”, Yelle Buffel (Bachelor Agro- en biotechnologie, KATHO Roeselare).

to kinin Pertry Ine

cyto molecular molecular

ysis contributes anal in-depth fascians cal locus

an fas ochemi the How Rhodococcus production: and bi

How the fas locus contributes to Rhodococcus fascians cytokinin production Ine Pertry ics technology net e Bio the requirements for of t and G n e illm f ers hnology in Sciences: ful ) l ec a t i reecke t 009 o r i e a (PhD V B p t 08-2 0 es lan c 2 Dr. M. Holst

: P r Dr. D. doctor rof. of yea P tor:

bmitted in u s or: ment University University

emic romo

epart

romot

he degree of Ghent D

Faculty of Scien Faculty Acad

Co-P

Thesis

P t